Read the Impact Assessment

Home , Beer in Austria, Beer in Poland, Beer in Portugal, Beer in Romania, Denatured alcohol

EUROPEAN COMMISSION

Brussels, 7.6.2018 SWD(2018) 259 final /2

This document corrects document SWD(2018) 259 final of 25.05.2018. The annexes were not included. The text shall read as follows:

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

Accompanying the document

Proposal for a Council Directive

amending Directive 92/83/EEC on the harmonization of the structures of excise duties on alcohol and alcoholic beverages

{COM(2018) 334 final} - {SEC(2018) 254 final} - {SWD(2018) 258 final}

EN EN

CONTENTS 1. Introduction and Context ...... 4 1.1. Introduction ...... 4 1.2. Scope for reforms ...... 5 2. What is the problem and why is it a problem? ...... 5 2.1. Introduction ...... 5 2.2. Scope of the problems...... 6 2.3. Problem 1 - Dysfunctions in the application of exemptions for denatured alcohol ...... 6 2.4. Problem 2 – Dysfunctions in the classification of certain alcoholic beverages ...... 10 2.5. Problem 3 – Dysfunctional application of reduced rates ...... 18 2.6. Problem 4 – Unclear provisions to measure the Plato degree of sweetened / flavoured beer23 2.7. Conclusion ...... 26 3. Why should the EU act? ...... 26 4. What should be achieved? ...... 27 4.1. General objectives ...... 27 4.2. Specific objectives ...... 28 5. What are the various options to achieve the objectives? ...... 28 5.1. Link between problems/drivers and options ...... 28 5.2. Dysfunctions in the application of exemptions for denatured alcohol ...... 31 5.3. Dysfunctions in the classification of certain alcoholic beverages ...... 33 5.4. Dysfunctional application of reduced rates ...... 37 5.5. Measurement of Plato degree for sweetened / flavoured beer ...... 39 6. What are the impacts of the different policy options and who will be affected? ...... 40 6.1. Dysfunctions in the application of exemptions for denatured alcohol ...... 41 6.2. Dysfunctions in the classification of certain alcoholic beverages ...... 42 6.3. Dysfunctional application of reduced rates ...... 46 6.4. Unclear provisions to measure of Plato degree for sweetened / flavoured beer ...... 48 7. How do the options compare? ...... 50 7.1. Dysfunctions in the application of exemptions for denatured alcohol ...... 50 7.2. Dysfunctions in the classification of certain alcoholic beverages ...... 53 7.3. Dysfunctional application of reduced rates ...... 58 7.4. Unclear provisions to measure Plato degree for sweetened / flavoured beer ...... 61 7.5. Summary of preferred package of options ...... 63 8. REFIT (simplification and improved efficiency) ...... 65 8.1. Context, methodology and constraints ...... 65 9. How would actual impacts be monitored and evaluated? ...... 68 10. ANNEXES ...... 70 ANNEX 1. Procedural Information ...... 70 1

ANNEX 2. Synopsis report – Stakeholder consultation ...... 77 ANNEX 3. Who is affected by the initiative and how ...... 90 ANNEX 4. Analytical Models used in preparing the Impact Assessment ...... 95 ANNEX 5. Council Conclusions (6 December 2016) ...... 98 ANNEX 6. Drivers of the dysfunctions in the application of exemptions for denatured alcohol 101 ANNEX 7. Classification of alcoholic beverages ...... 105 ANNEX 8. Drivers of the dysfunctions in the classification of certain alcoholic beverages 113 ANNEX 9. Tax, affordability and consumption ...... 116 ANNEX 9A Excerpt from Annex 6a of the external evaluation (by Ramboll) ...... 123 ANNEX 10. Reduced rates for other small producers ...... 132 ANNEX 11. Driver of the dysfunctional application of reduced rates ...... 142 ANNEX 12. Driver of the unclear provisions to measure the Plato degree of sweetened / flavoured beer 143 ANNEX 13. Methodology for measuring the Plato degree ...... 144 ANNEX 14. Country-specific analysis of impacts of policy options for measuring Plato degree in sweetened/flavoured beer ...... 147 ANNEX 15. Assessment of the impact on the market of the re-classification of certain products 153 ANNEX 16. Problems that will not be addressed within this impact assessment ...... 159 ANNEX 17. Comparison of impacts of alternative options ...... 160 ANNEX 18. Monitoring and Evaluation of Impacts ...... 168

ABBREVIATIONS

ABV Actual Alcoholic Strength by Volume AFC Alcohol as a Flavour-Carrier AWP Aromatised Wine Products BTI Binding Tariff Information CDA Completely Denatured Alcohol CJEU Court of Justice of the European Union CN Combined Nomenclature CNEN Combined Nomenclature Explanatory Note(s) EMCS Excise Movement and Control System EPC Excise Product Code ET Ethyl Alcohol EU European Union FPG Fiscalis Project Group hl Hectolitre IP Intermediate Products ITEG Indirect Tax Expert Group mn Million MS Member State(s) OFB Other Fermented Beverages OPC Open Public Consultation PDA Partially Denatured Alcohol REFIT Regulatory Fitness and Performance Programme SME Small and Medium-sized Enterprise TFEU Treaty on the Functioning of the European Union WHO World Health Organisation

1. INTRODUCTION AND CONTEXT

1.1. Introduction

The European Commission is committed to preventing trade distortions in the single market, ensuring fair competition between businesses, and reducing administrative burdens and compliance costs for businesses and tax administrations. The launch of the single market resulted in the abolition of tax controls at the borders between Member States (MS) and the adoption of common rules for excise products, including alcohol, to facilitate cross-border trade and to prevent competitive distortions. Excise duties for alcohol are regulated through two directives:  Directive 92/84/EEC1 on the approximation of the rates of excise duty on alcohol and alcoholic beverages sets out the minimum rates of excise duty on alcohol products.  Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages sets out the common rules on the structures of excise duty applied to alcohol and alcoholic beverages. This directive defines and classifies the different types of alcohol and alcoholic beverages, according to their characteristics, and provides a legal framework for reduced rates, exemptions and derogations in some sectors. In addition to these directives, Directive 2008/118/EC2 sets out the common provisions, which apply to all products subject to excise duties. This directive is currently under review in a separate proposal. Furthermore, businesses must adhere to other EU legislation, which regulates areas such as product definitions, labelling etc. Both alcohol Directives have failed to keep pace with developments including inflation. In 2006 in response to a request from Council, the Commission proposed to amend the minimum rates as set out in Directive 92/84/EEC. The proposal fell short of the necessary unanimity and was withdrawn in 2015 by the Commission. Since the adoption of Directive 92/83/EEC in 1992, the first and only evaluation of the Directive was in 2014. This Directive has not kept pace with the challenges and opportunities offered by new technologies and developments within the alcohol industry. The Directive was identified by the Commission for a retrospective evaluation under the Commission’s Regulatory Fitness and Performance Programme (REFIT). One of the objective was to identify weaknesses in the legislative environment caused by the Directive resulting in negative consequences for the stakeholders (e.g. obstacles to the functioning of the internal market, competitive disruptions, administrative and compliance costs, etc.) This impact assessment intends to ensure that the future proposal is cognisant of previous experiences and in particular identify any areas where the regulatory framework can be improved to bring benefits to businesses, MS and citizens. While a proposal to amend Directive 92/84/EEC may reduce the incentive for tax evasion and positively impact on public health, this impact assessment will not focus on this due to the limited support of stakeholders and the Commission's prior experience in proposing an amendment to this Directive. Furthermore this proposal will focus solely on requirements imposed by tax legislation and not sector / industry requirements.

1 Council Directive 92/84/EEC of 19 October 1992 on the approximation of the rates of excise duty on alcohol and alcoholic beverages, OJ L 316, 31.10.1992, p.29. 2 Council Directive 2008/118/EC of 16 December 2008 concerning the general arrangements for excise duty and repealing Directive 92/12/EEC, OJ L 9, 14.1.2009, p.12-30.

1.2. Scope for reforms

To support the REFIT evaluation, an independent study was carried out in 2014/2016 by a consortium led by Ramboll Management Consulting (hereinafter the ‘Ramboll Evaluation’).3 The recommendations and findings of the Ramboll Evaluation were taken into account in the Commission’s report submitted to the Council in October 20164. According to this Report, the Directive has proven to be effective and generally appropriate for the collection of excise duties. Nevertheless some problems have been identified and inefficiencies persist causing possible distortions of the internal market. The large variation in duty levels between MS5, which provides a strong incentive for tax evasion, and other weaknesses in the design of the tax necessitate the use of burdensome administrative procedures for both tax administrations and businesses. These increased administrative and compliance costs for businesses restrict the participation of small and medium-sized enterprises in intra-EU trade in alcohol and alcoholic beverages. In December 2016, Member States unanimously supported the call to review the Directive and the Council subsequently adopted Council Conclusions (see Annex 5), asking the Commission to carry out the necessary studies and consultation to submit a proposal for revision. In March 2017, the Inception Impact Assessment6 on a possible revision of the Directive was published, and laid down the problem areas to be assessed and a preliminary set of potential policy options. A grouping led by Economisti Associati s.r.l. (EA) and including the Centre for European Policy Studies (CEPS), CASE - Center for Social and Economic Research, wedoIT-solutions GmbH, and ECOPA undertook the assignment titled “Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages” (“the Study”). The Study analysed the scale of the problems identified in the Ramboll evaluation, assessed their evolution and assessed the impacts of possible options to address the problems identified.

2. WHAT IS THE PROBLEM AND WHY IS IT A PROBLEM?

2.1. Introduction

As noted above, the Ramboll Evaluation found that the Directive has proven to be effective and generally appropriate for the collection of excise duties. However some problems were identified and inefficiencies persist. These findings are evident by the frequent queries from MS and businesses, Indirect Tax Expert Group (ITEG) and Committee for Excise duties (ExComm) agendas, complaints against the Commission and the existence of the Fiscalis Project Group 013 on arrangements for taking forward the work on completely and partially denatured alcohol. The problems touch upon the following 4 areas: (i) Exemptions for denatured alcohol, (ii) Classification of certain alcoholic beverages, (iii) Reduced rates for small producers and low strength alcoholic beverages, and (iv) Measurement of Plato degree of sweetened/flavoured beer. The problem areas are very distinct from each other, which has an impact on the structuring of the analysis presented in this report. There is no uniform and homogenous market for alcohol and alcoholic beverages; the markets for specific beverages or other alcoholic products are generally not competing against each other, follow specific sectorial regulations and requirements, and exhibit specific problems. There are at the same time issues with the functioning of Directive 92/83/EEC which are of horizontal nature, such as the classification problems.

3 Ramboll Management Consulting, Coffey, Europe Economics, “Evaluation of Council Directive 92/83/EEC on the harmonisation of the structures of excise duties on alcohol and alcoholic beverages”, 2016 4 'Report from the Commission to the Council on the evaluation of Council Directive 92/83/EEC on the structures of excise duties on alcohol and alcoholic beverage', Brussels, 28.10.2016, COM (2016) 676 final. 5 Council Directive 92/84/EEC sets the minimum rates of excise, which is not within the scope of this document. 6 'Inception Impact Assessment on the Structures of excise duties on alcohol and alcoholic beverages', 01.03.2017

All of the problem areas require dedicated scope of analysis. The drivers behind each of the problems areas are problem-specific and so are, mostly, the consequences. As a result, also the objectives are drawn up in such a way that they correspond only to specific problems/drivers (see section 4). Acknowledging the complexity of the issues at stake and their analysis, Figure 2 offers an overview of the intervention logic behind the initiative, guiding the reader through the analysis. Furthermore, it should be noted that some of the problem areas impact specific stakeholders only, who have learned for the main to work within the current legislative framework. The general muted response can be attributed in part to the perceived risks of amending a Directive that is 25 years old, the risk of positive rates for some products and also to the fact that the problem may not impact the stakeholder in any event. Contrary to the apparent lack of interest of stakeholders, this reluctance cannot be generalised and Member States unanimously supported a proposal for an amendment to the Directive. 2.2. Scope of the problems

The Ramboll evaluation and the Study took a broad approach to the possible problems, identified through various sources, with the functioning of Directive 92/83/EEC. Follow-up analysis of both studies resulted in a conclusion that not all of the aspects of the problematic areas merited EU action. The excise duty exemption for private production of fermented beverages (i.e. beer, wine and other fermented beverages (OFB)) for home consumption, which was reviewed in both studies, will not be further considered in this impact assessment. The reasons behind this decision are explained in detail in Annex 16. The following problem tree outlines the problems, the drivers and their consequences of the problematic areas retained for further analysis in this impact assessment. Figure 1 –the problem tree

2.3. Problem 1 - Dysfunctions in the application of exemptions for denatured alcohol

Overall, the Study suggests that the EU regulatory framework for exempting denatured alcohol from excise duty works relatively well. The original objective behind the provisions – that of ensuring fair competition between businesses - was found to be still relevant. However, it is evident (inter alia from the frequent discussions within the ExComm, the ITEG and the Fiscalis 6

Project Group dating back to 2008) that the provisions in Art. 27 of the Directive concerning denatured alcohol are not phrased in a completely clear and unambiguous way, which has given rise to uncertainties and disputes, especially when denatured alcohol is moved across borders between MS whose interpretation of the applicable rules differ. The original intention of the provisions for exemption of denatured alcohol and in particular the differentiation between the exemption under Article 27.1 (a) and 27.1 (b) is no longer met under the current interpretation. Some of these uncertainties have non-negligible cost implications for producers and/or users of denatured alcohol, and can inhibit intra-EU trade in denatured alcohol.

Art. 27 stipulates that alcohol shall be exempted from excise duty if it has been denatured (i.e. the addition of certain substances to make it unfit for human consumption). It distinguishes between ‘completely’ denatured alcohol (CDA), for which there is a system of mutual recognition of national denaturing formulations to ensure it can be traded freely throughout the EU, and so- called ‘partially’ denatured alcohol (PDA), for which the exemption is conditional on its use for the manufacture of any product not destined for human consumption, and MS are free to define their own national procedures.

CDA is predominantly used for industrial use, whereas PDA is used for products not intended for human consumption but for which the rules on CDA are not suitable (i.e. because the intentionally strong smell of CDA means it cannot be used in perfumes or its tasting agents cannot be used with products that come into contact with the mouth, etc.). Examples of such products include cosmetics, perfumes, inks, screenwash and anti-freeze, detergents, paints and coatings, as well as biofuels, which account for the largest proportion by far.

The key drivers of the problem, which are discussed in Annex 6, are (i) an incomplete / inconsistent mutual recognition of CDA between the MS, (ii) divergent national approaches to PDA, (iii) divergent interpretations of certain terms related to PDA, and (iv) potential for fraudulent use of denatured alcohol. 2.3.1. Consequences: who is affected and how?

Member States authorities Fiscal fraud with denatured alcohol is estimated to result in lost tax revenues in the region of EUR 150-200 million per year across the EU (the bulk of which is in Central / Eastern MS). The World Health Organization (WHO) has published estimates that around 17% of all alcohol consumed in Europe in 2010 was unrecorded. The estimated proportion of unrecorded alcohol (based on data from the Commission's excise duty tables (EDT), 2016) ranges from as little as 3% (FR) to over 20% (RO). Box 1 illustrates the scale of the tax revenues lost in PL.

Box 1 - Estimating fraud with surrogate alcohol in Poland The WHO estimates the consumption of illicit alcohol in PL to be about 1.6 litres of pure ethanol pp/year (13% of total consumption). According to interviewees, the illicit alcohol is predominantly spirits (ethyl alcohol), its total legal consumption is about 120 million litres of pure ethanol/year. A project carried out in 2012 by the Polish Spirits Industry in cooperation with the Ministry of Finance found that, between 2009 and 2011, the majority of illicit spirits (7 out of a total of 12 million litres of pure alcohol/year) consumed in Poland were derived from decontaminated/purified industrial alcohol.7 Based on the current excise duty and exchange rates, this would be equivalent to just under EUR 95 million of excise duty lost per year (or approx. 6% of the total excise duty receipts from ethyl alcohol).8 Whether this is a realistic estimate depends on who is asked: while the authorities in PL estimate that the consumption of illicit alcohol has fallen to around 5% of the total recently (meaning this type of fraud is responsible for around EUR 50 million of lost revenue/year), some industry representatives reckon the

7 Based on OECD, ‘Illicit Trade: Converging Criminal Network. The size, impacts and drivers of illicit trade in alcohol’, 2016 8 Calculations based on data from the Commission’s excise duty tables (2016).

market share of illicit spirits in PL is closer to 20% (equalling approx. EUR 200 million per year of lost revenue). Source: EA, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017

A minority of MS also indicated they felt that the existence of various denaturing methods across the EU made it “particularly difficult” for their administration to monitor and control of the production and/or movement of denatured alcohol. This was mainly due to a lack of knowledge of the different denaturing formulations used by other MS, resulting in a burden on the time and resources available for analysis in the laboratories.

According to the CZ and PL authorities alcohol can account for between 25 – 50 % of their customs laboratories' workloads. Recently their laboratories have analysed several hundred samples of denatured alcohol. As a sample can be denatured using one of the many denaturing formulations, the list of ingredients to be checked can vary substantially in each sample making it impossible to establish chemical algorithms. However, only a minority of these samples contain (cleaned up) denatured alcohol of unknown origin. In these cases, testing these samples can reportedly be extremely time and labour intensive i.e. take several days and cost up to EUR 1 000 in staff time and materials. Assuming 500 samples are analysed each year and approximately 5 % of these are difficult cases, the estimated total annual cost for an administration is approximately (a maximum of) EUR 25 000.

Finally as described under the drivers, the different approaches to PDA lead to the legal uncertainty and legal proceedings, which have costs for both tax authorities and businesses. In the context of the Ramboll evaluation Member States highlighted the need for clear rules on the exemption of denatured alcohol. One MS noted that the "definitions of rules at this [EU] level is of utmost necessity, otherwise each MS will have its own system, according to its national interests, and that will only complicate matters."

Businesses The different procedures and regimes in each MS as regards CDA make cross-border trade more difficult and can create competitive advantages to some, particularly in the cross-border movement of CDA or imports from third countries. However, since extra-EU imports are subject to an import tariff of EUR 10.2/hl, and the value of all imports of denatured alcohol into the EU amounts to approx. EUR 20 million/year of which CDA accounts for a small fraction, no stakeholders consulted raised competition from third countries as a substantial concern either9. It therefore seems that it remains more a potential risk than a manifested distortion.

Nevertheless, there appears to have been a perceptible hindrance and cost associated with moving CDA cross-borders, as reported by OPC respondents. 9 (24) businesses indicated that they, and/or a company that they had done business or were in direct contact with, had incurred additional costs and burdens because alcohol recognised as CDA in one MS was not recognised as such in another MS, on one or more occasions. A further 8 respondents noted that alcohol was recognised as CDA after a delay. 7 respondents indicated that they or another company had chosen not to import/export CDA from/to another MS because of the risk it would not be recognised as such.

With regard to the administrative burden, in general, the main concern of the economic operators was linked to the specific requirements regarding supervision of production and movement in some MS that cannot be directly linked to the provisions of the Directive, and which represent these MS’ national-level response to their estimations of the risk of fraud.

9 Only 4 out of 21 OPC respondents indicated they or another company had chosen to purchase CDA from a third country, rather than from an MS, because it was subject to more lenient rules.

With regard to the different national approaches to PDA, while producers noted that intra EU trade in PDA is possible and does happen, the investment required (i.e. purchase of storage facilities, setting up and maintaining tax warehouses) is likely to prevent many. Furthermore the cost of understanding and complying with the applicable rules in the different MS hampers smaller businesses in cross-border trade. For example in the CZ the financial guarantee is approximately EUR 10/L of alcohol. Interviewees in PL noted that it was common for users of PDA to have a full-time member of staff dedicated entirely to ensuring compliance with the regulatory framework, and companies prefer to use CDA wherever possible to avoid the burdens associated with using PDA. Large specialised companies for whom alcohol is a key product often find it economical to make this investment (no detailed cost estimates were available).

In addition to the costs of complying with the supervisory regimes and providing the required information to national authorities, there can also be other operating costs that arise from the different procedures in different MS. For example, cosmetics companies in some MS have access to a much wider range of PDA formulations and production procedures (including in situ denaturation) than other MS. This can mean that the manufacture of certain products is possible to a higher standard and/or at lower costs in certain MS.

Multinational companies may be able to take advantage of such differences and locate the production of certain goods in the MS that offer the most favourable conditions as regards PDA formulations and related factors. However, there are also instances where multinationals companies in several MS claim to incur additional costs, as they need to adapt the formulations and production processes for otherwise identical products containing alcohol to the respective national PDA rules. The costs arising from the national regulatory frameworks vary significantly from MS to MS, from sector to sector, and even from company to company, and would therefore be very difficult to estimate comprehensively. In any case, the costs arising from these aspects are not attributable to the Directive, but to national implementing rules.

In response to the OPC, 57% of respondents indicated that they or a company they had contact with had incurred additional costs to understand the legal situation as regards the applicable rules and procedures for PDA when moved from / to another MS. 68% indicated that they had incurred additional costs / administrative burdens to ensure that PDA using a formulation accepted in one MS was also recognised as such in another MS. 39% of respondents had to pay excise duty on denatured alcohol because a MS did not recognise the procedure by which it was denatured in another MS and 48% of respondents chose not to import or export PDA due to risk it would not be accepted as PDA. No specific estimates or evidence was provided by the MS, which most likely stems from the fact that the administrations or companies do not keep such readily available statistics and disaggregating from other data is difficult. In the context of the Ramboll evaluation, one producer described a situation where a commitment had been made with a customer for the use of a specific denaturant. The authorities subsequently refused to authorise this formulation but the producer was contractually bound to produce the alcohol without an excise duty exemption.

With regards to the lack of clarity around the terms Art. 27 (1) (b) and the diverging interpretations of the term 'used for the manufacture of' the different interpretations by different MS, sometimes even by different customs offices within a given MS, of what does and does not constitute a finished product lead to legal uncertainty for businesses and costs (if the classification is challenged). Some businesses (47%) reported having experience of such situations, but were not able to specify costs, and stated the issue was eventually resolved to their satisfaction (in one case via a BTI). In MS where the exemption is not applied, the businesses may have to incur additional costs for purchasing and storing CDA in addition to PDA. The majority of stakeholders (76%) involved in the production or end-use of industrial alcohol stated that they have encountered issues with different interpretations.

Consumers/Citizens Apart from the lost tax revenue, the resulting reduced funding for public services and other negative effects of criminal activity on society, the other main concern in relation to the effects of this fraud is public health. There is at least one known recent case in the UK where anti-freeze containing denatured alcohol seems to have been used to manufacture illicit vodka. The consumption of denatured alcohol is also evident in LT, where it is commonly known that mouthwash is sold to individuals for consumption as alcoholic beverages. The Polish National Health Fund data show an average of around 200 hospital admissions, and around 50 deaths due to glycol (alcohol) poisoning per year in PL.

2.3.2. How will the problem evolve (baseline scenario)?

Approx. EUR 3-3.5 billion worth of denatured alcohol is used annually in the EU for a variety of industrial purposes. It is estimated that more than 95% of the total consumption is PDA, although CDA accounts for a significant share of the market in certain MS and sectors.

With the adoption of Regulation 2017/2236 and the entry into force of the new list of CDA formulations, 25 MS recognise the Eurodenaturant as the only denaturing formulation, with only 3 MS (CZ, SE, UK) recognising different concentrations of the same ingredients. In addition, from 2019, when the authorisation of the remaining FI formulation expires, only 2 MS (CZ, EL) will still be using national formulations containing different denaturants. This greatly reduces, but does not completely eliminate, the scope for problems arising from the manifestly unclear rules on recognition of CDA formulations stipulated in Art. 27(1)(a) of the Directive. Still, it should be noted that the Directive in its current form allows MS to re-introduce national CDA formulations, if they wish to in the future. While this seems unlikely, it cannot be ruled out, especially if issues with the Eurodenaturant were to come to light.

In any case, if one considers that a system is only as strong as its weakest link (as CDA can circulate freely across the EU, and fraudsters would tend to use the “weakest” formulation available), then the fact that many MS have replaced their national formulations with the Eurodenaturant should reduce the risk of fraud with CDA overall. It is impossible to predict if this will result in a reduction of fraudulent activity or in a displacement of fraud towards PDA.

The proliferation of national approaches to PDA will continue. It could possibly intensify for biofuels, which accounts for the largest proportion of PDA as the future market evolution of biofuels is dependent on the direction of renewable energy policy in Europe. No changes are expected in relation to the divergent interpretations related to PDA and the uncertainty for cross- border trade will continue.

2.4. Problem 2 – Dysfunctions in the classification of certain alcoholic beverages

2.4.1. The problem and its EU dimension

Alcoholic beverages are defined and categorised at multiple levels and for different purposes. These different layers only partly coincide and this lack of coherence seems the single most significant cause of all classification issues. The excise duty classification is determined by the five harmonised tax categories established in the Directive, which are defined primarily with reference to the Combined Nomenclature (CN) headings. See Annex 7 for details.

The classification dysfunctions, which can be subdivided into two key areas which exhibit distinct characteristics (and drivers) while sharing most of the adverse effects: (i) interaction between fiscal and customs classification and (ii) definition and classification of certain non- standard products not explicitly, or imprecisely, foreseen in the Directive. The Ramboll

evaluation recalled approximately 70 different cases of products “difficult to classify”, spanning a majority of MS.

Interaction between fiscal and customs classification

The Directive defines the categories of alcoholic products subject to harmonised excise duty in accordance with their customs classification. The correspondence between the fiscal categories and the CN codes is however not straightforward. Within the EU, classification uncertainties have lead to disparities of treatment across MS and between similar products, due to different criteria used to determine the essential fermented character of certain beverages. The level of the legal uncertainty that may derive from the above classification issues is connected primarily to the specificities of national markets, and the classification rules adopted.

Under the current system the customs classification determines the excise duty category. Once a beverage is classified as CN 2208 it can be taxed only under Art. 20 (Ethyl alcohol), while if classified as CN 2206 it may fall under Art. 12 (OFB) or 17 (IP) depending on its strength, but not under Art. 20 (unless it exceeds 22% vol., but there are no actual market incidences). Since the excise duty classification follows the CN classification, administrations have limited room for manoeuvre in applying the category that they consider appropriate for products that has a CN code they disagree with. In principle, tax administrations might challenge questionable CN coding decisions, but when these are covered by a Binding Tariff Information (BTI) issued in another MS they generally opt to avoid disputes. The consequence is that similar products may end up being subject to different excise categories depending on the country of origin.

The magnitude of the problem is reflected in the number of Court of Justice of the EU ("CJEU") rulings on the classification of alcoholic beverages, which captured some instances of disagreements and disputes over the classification of products that took place in the various MS. In fact, especially where the matter is in the remit of customs offices instead of tax offices, the disputes are reportedly settled through alternative methods: when a misclassification is detected, the competent administration imposes the payment of a certain amount of tax arrears (with/without a fine) to the responsible entity. Businesses prefer this procedure rather than opening a legal case, since it is faster, it often envisages the possibility of negotiations, and it does not imply public disclosure so the potential reputational effects are minimised. However, for this very reason precise figures on the frequency of administrative cases are not available.

The landmark rulings of CJEU (see box 2) established the possibility of classifying dubious products and gave MS a tool to tackle opportunistic practices. On the one hand, the CJEU rulings effectively indicated how to interpret the old rules vis-à-vis new products, but on the other hand the selection criteria remained somehow subjective (taste, smell, appearance) or indefinite (no specific thresholds or methods to determine the prevalent origin of the alcohol used). Therefore the risk of disparities in the application of these criteria across national administrations persist, and the need for objective classification rules has possibly become even more pressing.

Box 2 – Summary of CJEU landmark cases on the classification of alcoholic beverages Case C-150/08 (‘Siebrand) regarded alcoholic beverages – in specific the three beverages ‘Pina Colada’, ‘Whiskey Cream’ and ‘Apfel Cocktail’ – with a cider base to which distilled alcohol, water, sugar syrup and various additives had been added. The question was if these beverages may maintain the CN 2206 code – due to their cider base – or should be classified under CN 2208 as established by the Dutch customs. The Court ruled that when a fermented beverage loses the taste, smell and/or appearance of a beverage produced from a particular fruit or natural product, due to the above mentioned additions, it no longer falls under CN 2206, but CN 2208 applies.

Case C-196/10 (‘Paderborner Brauerei’) concerned the fermented beverage ‘Salitos Ice’ and its ‘malt beer base’. The ‘malt beer base’ was produced from brewed beer with an alcoholic strength by volume of approx. 14%, which was clarified and then processed with ultrafiltration techniques. The base obtained was then employed for the production of a light beer-based mixed drink. The question was if such a product had to be classified under CN 11

2203 or 2208. The Court ruling established that ‘a liquid described as a “malt beer base”, with an alcoholic strength by volume of 14%, obtained from brewed beer which has been clarified and then subjected to ultrafiltration, by which the concentration of ingredients was reduced, must be classified under heading 2208”.

The joined cases C-532/14 and C-533/14 ( ‘Toorank’) tackled the fermented beverage called ‘Ferm Fruit’ and a range of beverages with a ‘Ferm Fruit’ base to which other ingredients were added. Ferm Fruit was prepared using an alcohol resulting from the fermentation of fruit, which was then purified through ultrafiltration so that its smell, colour and taste resulted neutral. The question was if ‘Ferm Fruit’ (Question 1) and ‘Ferm Fruit-based beverages’ (Questions 2&3) had to be classified under CN 2206 or CN 2208. The CJEU ruled that ‘a beverage, such as Ferm Fruit, which is obtained through fermentation of an apple concentrate and is designed to be consumed either undiluted or as a base in other beverages, being neutral in terms of colour, smell and taste as a result of purification (including ultrafiltration) and having an alcoholic strength by volume, without the addition of distilled alcohol, of 16% falls under heading 2208 of that nomenclature’.

Source: EA, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017 Definition and classification of certain products

Harmonised EU definitions of some alcoholic products exist. In the case of spirits, this harmonised definition was developed to ensure, amongst other things, a systematic approach to spirits, to prevent the misuse of the terms and to protect the reputation of EU spirits10. However there is no harmonised definition of cider, perry and fruit wines in the Directive or in other EU legislation. Both in the CN and excise classifications, the OFB definition is less strict than for other alcoholic beverages. This reflects at the same time the heterogeneity of the products comprised (e.g. cider, perry, mead, other fruit-wines, and mixes), the variety of national production practices (‘cider’ designates products with marked differences across MS) or commercial designation of these products (e.g. malt-based alcopops, 'wine-coolers', un-hopped flavoured beer, cider and fruit wine based refreshers, generic low-strength pre-mixes, certain cream liquors and other flavoured liquors, etc.) and the related absence of harmonised sectoral definition and rules that to the contrary exist for wine and spirits.

Borderline products have been introduced to the market with the specific aim of being classified in a product category with a lower excise duty rate compared to competing products. Tax differentials vary and high differentials can be observed in MS with a zero rate on OFB.

A report11 shows that certain products that in PL are classified as spirits are very similar to other products that other MS classify as OFB. 65% respondents experience frequent classification uncertainties and disputes within the pre-mixed drinks product group. 62% said uncertainties and disputes frequently occur with the category fermented alcohol pushed to 15-21% actual alcoholic strength by volume (ABV) industrially, bottled and sold to look like its equivalent spirit, on which a higher excise duty is due.

This favourable tax treatment, combined with a certain flexibility of the criteria used to define this category, provided in the past an incentive for the development of various new products, based on novel production techniques, arguably designed to take advantage of the OFB tax category for competition purposes. In the absence of a harmonised definition, a number of MS have adopted national ad hoc measures for the tax treatment of OFB.

10 Regulation (EC) No 110/2008 of the European Parliament and of the Council of 15 January 2008 on the definition, description, presentation, labelling and the protection of geographical indications of spirit drinks and repealing Council Regulation (EEC) No 1576/89, OJ L 039 13.2.2008, p. 16. 11 Report prepared for the Polish Council of Wine by Parulski & Wspolnicy, 'Tariff and Excise Tax Classification of Fermented Beverages – Issues of Concerns', September 2016.

A final element of the classification issue relates to the disparities of classification of certain flavoured wine and OFB to which minimal amounts of alcohol which are not from fermented origin are added as a Flavour Carrier (AFC) or for other functional purposes. According to the Directive, the alcohol contained in a product should be of “entirely fermented origin” in order to be classified as wine or OFB. However, no clarity is provided by the Directive for products containing both alcohol from ''entirely fermented origin'' and ''non-entirely fermented origin'', and the some disparities may arise, regarding aromatised wine products (AWP) or flavoured OFBs.

Similarly, the CN 2206 heading admits products not entirely of fermented origin12, but the permitted amount is not specified13, and the jurisprudence in this area (see box 2) did not establish any straightforward criteria. As a result, various MS have already adopted non harmonised provisions establishing a margin of tolerance for products containing AFC by either (i) adopting a flexible approach to functional alcohol added, or (ii) setting specific maximum limits (in ABV terms) to the amount of AFC that can be added to a fermented beverage before the tax category changes (typically 1.2% vol).

To the extent the tax differential between Art. 12 and 17 and between Art. 17 and 20 is high, there remains incentives for certain businesses to exploit this ambiguity. It is difficult to accurately quantify the size of this market, however 2017 estimates are in the region of 850 billion litres. Approx. 550 billion litres of which are flavoured beer, which does not pose classification issues as Regulation 1967/200514 addressed this issue. However this is a growing area, although moderately, and the risk of abuse may become more relevant in the future.

The key drivers of the classification problem are discussed in detail in Annex 8.

2.4.2. Consequences: who is affected and how?

As mentioned above, the Ramboll evaluation identified approximately 70 different cases of products “difficult to classify”, across most of the MS. While the consequences surrounding each case are unique (some were resolved swiftly following a few exchanges between the tax administration and the economic operator in question, while others became the subject of lengthy court cases spanning several years), it is clear that all the cases have resulted in additional administrative burdens for the tax administrations (who had to dedicate additional resources to enforce their view of the correct classification) and compliance costs for economic operators (who needed to undertake similar actions to defend their position against either the tax administration or a competitor).

An important outcome revealed in relation to the situations documented was litigation costs. Disputes between tax administrations and operators were likely to be taken to court, resulting in significant costs both for the administration and for the economic operators if the financial risk at stake was considerable. Additionally, litigation resulted in significant costs for economic operators seeking to correct the perceived unfair competition presented by “difficult to classify”

12 The explanatory notes and classification opinions adopted by the Harmonised System Committee relating to Heading 2206 states: “All these beverages may be either naturally sparkling or artificially charged with carbon dioxide. They remain classified under this heading even when fortified with added alcohol or when their alcohol content has been increased by further fermentation, provided that they retain the character of products classified under this heading.” 13 When goods are prima facie classifiable under two or more headings, the CN rules require that classification is effected as follows: “mixtures, composite goods consisting of different materials or made up of different components, and goods put up in sets for retail sale, (…), shall be classified as if they consisted of the material or component which gives them their essential character, in so far as this criterion is applicable”. 14 Commission Regulation (EC) No 1967/2005 of 1 December 2005 concerning the classification of certain goods in the Combined Nomenclature, OJ L 316, 2.12.2005, p. 7–9

products. This has been particularly observed in cases when high-strength mixtures emulated or directly competed with spirits or intermediate products which were taxed at a higher rate.

Member States authorities As far as the disparities between customs and excise duties are concerned, businesses may be tempted to request a classification in jurisdictions where it is more likely to obtain a more favourable (tax wise) classification, in order to get competitive advantages across all EU national markets. Reportedly, there have been cases of ‘BTI shopping’, i.e. demands submitted in MS where a favourable classification was considered more probable. However, the BTI rules and practices seem to be changing: BTI shopping has become less feasible, and some customs release BTI decisions only to products for export. Still, BTIs are not exempt from disputes, although concrete cases are rare, and customs authorities rarely challenge a BTI issued in another MS.

Besides the risk and the costs of disputes, the lack of clear criteria and parameters for certain ‘borderline’ products makes the process complex, long, and unpredictable for all involved. Although it concerns formally the customs classification, it is the consequential excise duty categorisation that is primarily at stake, so the administrative burden caused by CN classification should be considered as directly related to the functioning of the excise duty system. Burdens and costs related to these uncertainties for administrations and businesses vary considerably between MS. The Study estimates the costs at EUR 1-1.5 million/year at EU level.

The bulk of the extra burden is borne by national authorities. Eleven MS consulted in the context of the Ramboll evaluation agreed or strongly agreed that the difficulties encountered with the classification of alcohol and alcoholic beverages were leading to increased administrative costs. These costs relate primarily to the additional efforts required to deal with complex classification cases, including laboratory tests and the extra labour to manage the dossier and liaise with the applicant. Unfortunately, none of the eleven MS were able to specify precisely to what extent their administrative costs were greater than they would have been otherwise. As for anecdotal evidence, French authorities reported that 'the dispute on the classification with the producer of a product of fermented base which has been elaborated to resemble distilled alcohol requires nine employees of the tax and customs authorities to be involved'.

To cope with the mounting number of 'borderline products' various MS have established ad hoc expert groups responsible for defining detailed classification rules and procedures and ensuring consistency in their tax treatment. Typically, these groups operate at the central level, collating the difficult cases that cannot be solved by regional customs offices. An intensification of the collaboration and exchanges between customs authorities at EU and international level aimed at resolving the uncertainties in the interpretation of the subjective criteria concerning certain CN 2206 products, which also results in costs for MS has also been reported. Unfortunately the customs administrations interviewed were not in the position to estimate the frequency of problematic cases, and the administrative burden attributable to these dossiers.

The existence of tax incentives having a product classified within one excise category over another has resulted in the development and marketing of products which seek to comply with the requirements of a more beneficial tax category while arguably (i.e. in the opinion of MS tax administrations and some competitors) circumventing the intention of the legislator of what should fall within the more favourable category. These manufacturers are exploiting the uncertainties and this is depriving MS of tax revenues. However estimates of foregone tax revenues are highly speculative and it is not feasible to determine precisely what share of these products have been developed purely for tax optimisation purposes, or what is the importance of an advantageous tax classification vis-à-vis other factors. Annex 9A presents the results of case studies relating to the classification issues with reported examples of specific products. In a nutshell, depending on the characteristics of the products (e.g. the alcohol content), the

determined CN classification, the country in which it is being sold and other individual variables of each case, the differences in applicable excises duties can vary between:

 7.48 EUR/HL to 89.7 EUR/HL of finished product for ready-to-drink products (e.g. alcopops);  79.55 EUR/HL (a 10-12% ABV, “Irish cream” type product in the UK) to 256.864 EUR/HL (a cleaned up fermented alcohol at 14-15% with sugar, aroma, acidifier, colouring and fizz in France) of finished product for medium strength fermented beverages, and;  200.00/ HL (a 21% ABV, fermented beverage in PT) to 331.40 EUR/HL (a 22% special fermentation of 'made wine' decolourised and flavour stripped and then sold in Vodka style packaging in the UK) of finished product.

The lack of a harmonised approach for beverages containing AFC across MS could lead to adverse impacts on internal market functioning and tax revenues. In the absence of clear regulatory statue for AFC these products may be subject to classification disparities as well.

Finally the lack of a separate EPC for OFB is an issue for market monitoring and control purposes due to the lack of accurate data. This could result in the incorrect calculation of excise duty due and the associated financial guarantee required for intra community movements, resulting in disputes (and costs) between tax authorities and businesses.

Businesses The Ramboll evaluation concluded that the classification of most alcoholic beverages from an excise perspective was generally seen as straightforward and results in no administrative burden linked to the application of the obligations inscribed in the legislation. At the same time, the stakeholders pointed out that issues surrounding the "difficult to classify" products do however result in increased costs for all the stakeholders concerned; the high costs identified were the result of the complications and disputes arising from situations in which the stakeholders disagree on the correct interpretation of the provisions of the Directive. Nearly 30% of the economic operators consulted in the context of the evaluation reported that they had had difficulties with the assignment of alcohol and alcoholic beverages to the categories of the Directive. Difficulties were noted in all sectors but the beer sector indicated that these difficulties had led to increased administrative costs.

The costs implied for each organisation varies significantly depending on the evolution of a given case, the economic importance of the disputes, the willingness of the parties to settle the matter via the judicial system, etc.15 A representative of a trade association in the area of spirit producers indicated anecdotally that a court dispute over the classification of a product of fermented base with added ethyl alcohol lasted for four years. In fact, only five out of 43 trade associations responding to this question did not report that their administrative costs had increased due to classification problems.

This lack of clarity and legal uncertainties resulted in numerous CJEU cases in the past and high costs for businesses. While the number of cases reduced since the judgment of the CJEU in case C-150/08, there are continued disagreements. However due to the high costs borne by certain businesses, which saw their turnover halved and in some case almost caused their bankruptcy, there is limited appetite for more litigation. In fact, businesses have become more risk wary.

15 Precise monetary quantification of the expected cost has not been possible due to the varied nature of the cases reported according to the research conducted during the evaluation.

Launching new products in the absence of formal classifications may result in increased administrative burden and costs for businesses and delays in getting new products onto the market. According to stakeholders this is a serious issue as other countries' classifications, including third countries, are seldom the subject of legal challenges, thus creating competitive advantages for these businesses.

Misclassifications of products may result in higher excise duties for businesses and as a result higher financial guarantees may also be imposed. This may result in substitution effects if the higher excise duties are passed on in full to the consumers. There are conflicting data regarding substitution effects, however the introduction of the 'alcopop' tax in Germany is a classic example of how taxes can have a profound impact on substitution.

Box 3 – Possible substitution effects induced by the introduction of the ‘alcopop tax’ in Germany Useful insights on substitution effects between different alcoholic products can be drawn from the review of the consumption trend of alcoholic beverages in Germany between 2000 and 2007. In the first three years of years 2000s, mixed drinks grew in popularity and their consumption recorded an impressive growth (about 78% per year, on average), which partly offset the decline in the volumes consumed of beer and spirits. After the introduction of the alcopop tax in July 2004, consumers and the market responded negatively, and a major decline in consumption was recorded – i.e. amounting to some 50% per year between 2004 and 2006. Looking at the trend in consumption of other beverages, it seems that some previous drinkers of mixed drinks switched to beer as indicated by the slowing down of its declining rate. The existence of a similar substitution effect has been confirmed by a 2010 study to assess the effects of the alcopops tax on alcohol consumption and beverage preference among adolescents in Germany.16 Based on 2003 and 2007 data from the cross-sectional survey of the European School Survey Project on Alcohol and other Drugs, the study confirmed a partial substitution of alcopops by spirits and beer among 12–17-year-olds. Source: EA, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017

In addition to the quantifiable difference in terms of applicable excise duty as explained above, economic operators interviewed in the context of the case study on classification issues conducted under the Ramboll evaluation, have reported barriers to conducting business across the EU resulting from uncertainty with respect to the treatment of their product (i.e. being treated as W200/2206 in the home country, but considered S200/2208 in other MS). Another negative consequence concerned unfair competition aspects of the internal market; according to economic operators reporting examples of such products, the existence of this classification issue affects competition in two different ways:

 firstly, it places producers of similar products which are entirely from alcohol of distilled origin (which compete on the same market) at a severe competitive disadvantage (see above the difference in duty levels);  secondly, it undermines the excise category itself by allowing products to deliberately benefit from taxation at the same level as 'clear-cut' products whose protection the category itself was supposed to benefit.

Consumers/Citizens The relationship between tax, affordability and consumption at systemic level is in research systematically confirmed weak (see Annex 9). Therefore, the overall impact on per capita consumption of alcohol possibly caused by the tax-induced substitution between ‘standard’ and

16 Muller S, Piontek D, Pabst A, Baumeister SE, Kraus L., Changes in alcohol consumption and beverage preference among adolescents after the introduction of the alcopops tax in Germany. Addiction 2010; 105:1205–13.

‘borderline’ products is considered to be of modest magnitude. This does not evidently deny the existence of problems linked to the consumption of certain alcoholic beverages by certain socioeconomic segments of the population, which have been tackled inter alia through ad hoc national taxes, which was also confirmed by the public health stakeholders interviewed.

The lack of a direct correlation between tax policies and per capita consumption seems intuitively confirmed also by noting that the decline in the total alcohol consumption registered by WHO – Global Information System on Alcohol and Health (GISAH) was the strongest for spirits (-2.11% in seven years), but in almost the same period the registered increased in excise duty level was the lowest for this category (+ 2.4%).

Nevertheless, borderline products which enjoy a favourable tax classification may appeal to young people or vulnerable social categories with limited disposable income. Increase in the development of borderline products may result in increased consumption and overall create negative public health impacts.

2.4.3. How will the problem evolve (baseline scenario)?

Harmonised classifications of alcoholic beverages are of utmost importance for intra EU and international trade. In 2016, the value of alcoholic beverages exports and imports from/to the EU amounted to EUR 24 billion and 4.5 billion respectively. Classification uncertainties of alcoholic beverages may cause barriers to trade, market disruptions and enforcement problems.

The Study supporting this impact assessment suggests that the dimension of the categories containing borderline products are limited in volume terms. The mixed drink category amounts to an estimated 78 million litres that is approximately 6% of the ‘fiscal’ OFB category. In a micro perspective, mixed drinks with a fermented base may (and did) represent a cheaper alternative to spirit-based mixed drinks, thanks to the more favourable tax treatment, thus posing a potential competition issue.

Classification uncertainties and disputes are becoming less frequent due to the high litigations costs borne by certain businesses. As a result businesses including both brand owners, wholesalers and distributors have become more risk-wary towards the placement on the market of new products if not clearly identified. However the disparities persist due to the subjective rulings of the CJEU and new technological developments will continue to create uncertainties if the scope of the OFB category is not clarified.

Businesses are now using alternative methods, for which precise figures are not available. Some national customs have adopted rules and procedures to effectively operationalise these criteria. In FR, a specific platform, i.e. Soprano, has been established to this end. The platform allowed authorised businesses to submit classification dossiers to obtain a preventive opinion in a faster way. The pilot initiative was launched in 2017 and its use at the moment is voluntary, but if successful it might become the standard procedure for the submission of applications. In addition to preventing disputes, the expected benefits of Soprano also include a reduced length of procedures so a shorter ‘time-to-market’ for enterprises. Nonetheless, as the national approaches are non-harmonised at EU-level there remains the risk of different/incoherent legal interpretations and ensuing disputes, as well as incentives to continue to develop products exploiting these classification uncertainties.

Competitive advantages will persist for businesses who obtain a favourable tax classification, which will encourage 'classification shopping'. The nature of ‘borderline’ products is different across markets since it relates to specific consumer preferences and opportunities, but in general the problematic area seems to increasingly focus on fermented bases having undergone some form of concentration and/or cleaning, both traded as such or used in final beverages. Cases were 17

reported of products stored in the producers’ tax warehouses as CN 2208, then dispatched to another country as CN 2206; beverages moved in a bordering country, re-bottled and re- imported, with a more favourable classification; trade of entirely fermented bases with ABV of 22% coded as CN 2206 etc. In this respect, the products covered by the CJEU rulings are no longer the core of classification uncertainties and issue, but other new challenges are seemingly emerging.

The unclear application of 'entirely fermented origin' and the absence of a separate EPC may cause some market distortions and monitoring / control issues for some products. With respect to future trends, two considerations apply: (i) an increasing number of MS have adopted a flexible approach to AFC, possibly in connection with the EU-level legislation. This trend is likely to continue, since MS that have not set explicit threshold for AFC are reportedly inclined to maintain margins of tolerance in the classification of these products. So disparities of treatment are progressively less likely; (ii) on the other hand, the market size of these products is growing, although moderately, so the risk of abuses may become more relevant in the future.

2.5. Problem 3 – Dysfunctional application of reduced rates

2.5.1. The problem and its EU dimension

The scope and application of reduced rates to some alcoholic beverages is a multi-faceted problem, which could be sub-divided into more specific aspects. Whereas the Studies and the stakeholders consulted, globally consent that the reduced rates framework is working, there are issues that are acute to a specific industry or to a specific aspect of the legal framework.

Unequal treatment of producers of alcoholic beverages

The first sub-problem within the application of reduced rates evolves from the unequal treatment of producers of alcoholic beverages which can lead to market distortions. The Directive allows MS to grant reduced excise rates to small producers of beer (Art. 4) and ethyl alcohol (Art. 22) only; small producers of wine, OFB (including cider and perry) and IP are not subject to this provision. Even if the MS wanted to correct this imbalance, the Directive effectively prevents them from doing so. Ireland and the UK highlighted this unfair discrimination in the Ramboll evaluation. Detailed analysis of reduced rates and special schemes applied to all sectors of alcoholic beverages is included in Annex 10 with key aspects recapitulated here below.

When it comes to cider (and perry), in most countries, cider makers are not intermingled in complex relationships, and small cider producers make cider themselves, rather than providing products to larger companies. In terms of market structure, micro and small cider makers represent the vast majority of the population (96% in the UK, 99% in FR, 93% in IE), but a small share of the market.

The fortified wines industry includes growers, producers of the base wine and fortifiers. The vast majority of growers do not produce the end product. The number of small producers within this industry, who would be affected by the application of reduced rates is small.

For wine producers in MS applying a zero or near zero excise duty rate to wine, the introduction of reduced rates would bring no tax advantage to small producers and therefore the relative competitive position of drinks would not change. This is not the case in MS applying a positive excise duty rate. However in the view of stakeholders, the introduction of reduced rates for small producers of wine could result in the subsequent removal of the zero rate, an outcome which would negatively affect all businesses, both large and small.

Taking account of these factors, this impact assessment will focus on small cider makers only. For the sake of transparency and completeness the analysis of reduced rates for small wine producers and fortified wines together with options is presented in Annex 10.

Legal uncertainty

The second sub-problem area of the application of reduced rates concerns the lack of clarity of the current provision and the legal uncertainty thus created for the markets. The granting of reduced rates to small producers is conditional upon their independence in legal and economic terms from any other brewery and no operations under license. However, the Directive does not define the term 'legally and economically independent' and this has resulted in businesses consulting other EU law to resolve this17. With respect to beer brewed under licence, the issue has been largely resolved by existing guidelines and clarifications. Despite this, conflicts on the term 'legally and economically independent' between producers and authorities persist, which require legal proceedings, rulings and therefore litigation costs for both parties.

Box 4 – CJEU case C- 285/14: Brasserie Bouquet SA (FR)

Brasserie Bouquet operates a restaurant in which it sells beer it has brewed itself. It entered a membership contract with ICO 3B SARL, which authorised Brasserie Bouquet to use the trademarks, the commercial designation "Les 3 Brasseurs" and to receive ICO 3B SARL's know-how. In exchange Brasserie Bouquet paid an entrance fee and was required to exclusively obtain certain products from ICO 3B SARL.

Brasserie Bouquet considered it satisfied the conditions of the small brewery relief. The FR authorities challenged the application of the reduced rate that Brasserie Bouquet paid. The CJEU ruled that for the purpose of applying the reduced rate on beer the condition laid down in Art. 4(2) of the Directive according to which a brewery must not operate under licence, is not met if the brewery concerned makes its beer in accordance with an agreement pursuant to which it is authorised to use the trademarks and production process of a third party. The UK businesses consulted confirmed that 'contract brewing' may still have a certain degree of subjectivity regarding whether a contract breaches the independence of each counterpart or not. French stakeholders reported that this issue should have been settled by a Customs Memorandum, but this has led to different interpretations by local customs offices.

In terms of the cross-border functioning of the reduced rates for small brewers, MS report implementation problems, as customs authorities in the country where the product is released for consumption need to check the status of the brewer. UK authorities consider this to be a ‘self- declaration’ scheme, so that controls on businesses claiming the status of ‘small producer’ are risk based. In FR, a small brewer must make a one-off submission of a set of company documents.

In case of disputes, the customs authority in the MS of destination may submit a request for information to the customs authority in the country of origin to verify the status of a small brewer. This verification may be problematic in the case of businesses based in a non-EU country. However, most of the customs authorities interviewed do require a certificate from the brewers or their distributors, issued or stamped by the home country customs authorities. Businesses interviewed confirmed that, when moving products across borders, the local distributor may ask for such a certificate, but this does not always happen. A problem arises when (i) a small brewer established in a MS not requiring the certificate and not issuing the certificate to domestic manufacturers intends to enter the market of a MS requiring such a certificate; (ii) or when the MS of destination does not automatically recognise the status granted by the country of origin. FR allegedly does not accept self-certification and does not always recognise checks performed by

17 Commission Recommendation of 6 May 2003 concerning the definition of micro, small and medium-sized enterprises, C(2003) 1422, 6.5.2003 provides an explanation of when two companies should be considered partners or linked. It does not provide an explanation on 'brewing under licence' or 'contract brewing'.

the Belgian customs authority and this is affecting a significant number of Belgian producers. In this regard, BE noted during the Ramboll evaluation that at an administrative level there are a lot of problems regarding interpretation in order to determine the status of a 'small independent brewery'.

Even though the reduced excise rates for small breweries are estimated to cover only 5% of production, it is estimated that 95% of active breweries are covered by this relief. The problems described above may be of limited scope today. However with the continued increase in the number of small breweries and their growth into larger businesses, it is likely that cross-border trade flow will increase and the commercial relationships will become more complex. As a result these uncertainties are likely to evolve into bigger issues in the future.

The functioning of reduced rates for distilleries meets the same obstacle of an unclear definition of 'legally and economically independent' businesses. The rationale of this relief is to protect and preserve the traditional distilling culture. The distilleries benefiting from the reduction are the ones likely to work on an occasional basis, e.g. after fruit harvesting of grape pressing, selling their products, for the very local market. The threshold was therefore set much lower, making the commercial viability of such a scale of production extremely limited.

During stakeholder consultation most of producers in the ethyl alcohol industry expressed a negative opinion on the current threshold. While all considered that it was not fit for purpose the reasons differed. Most stakeholders showed limited, if any, interest in a revision of the threshold and most authorities expressed no intention to implement an amended provision at national level. Taking account of this, this impact assessment will not focus on this problem area.

Irrelevant and incoherent alcoholic strength thresholds for some product categories

The final problematic area of application of reduced rates relates to low strength alcoholic beverages. Art. 5, 9, 13, 18, and 22 of the Directive allow MS to apply reduced rates on low- strength alcoholic beverages, but the Directive is silent on the targets or objectives of these provisions. More specifically, it is not clear whether the option for reducing rates represents a tool to: i) tailor national taxation policies; ii) pursue objectives of industrial and agricultural policy; iii) incentivise product innovation; and/or iv) achieve health policy objectives. This is not generally perceived as an obstacle to its uptake in MS who are contented with the flexibility offered under the arrangements allowing them to pursue their own priorities and adapt the structure of the excise duty on alcohol to national needs. However the alcoholic strength thresholds to apply reduced rates are set at levels that are largely irrelevant for some product categories, while applicable to the entire market for other products. For example:

The current thresholds for wine (8.5% vol.), IP (15% vol.) and ethyl alcohol (10% vol.), do not Wine, to reflect the features of products included in these categories. Very few products in these intermediate categories could fall below the threshold. In most cases, to comply with product definitions products, ethyl spelled out in EU law, such products must have an alcohol content above the maximum alcohol thresholds set by the Directive.

The current threshold allows the application of reduced rates mainly to radler and very few Beer other products. It is too low to provide any tangible incentive for brewers to be innovative and create new low-strength products or for consumers to drink low strength beer.

The current threshold set for OFB (8.5% vol.), which covers almost the entire market for cider OFB and perry and a portion of the market for fruit wine, appears to pose no policy problem

It should be recalled that the present impact assessment does not touch upon the duty rates of excisable products but the excise duty structure. What is discussed in the present impact assessment is therefore not the levels of reduced rates for different categories of alcoholic beverages, but rather what products they may be applied to and under which conditions, including 20

thresholds. The concerns presented here do not relate to the functioning of the internal market as such – which is deemed to be functioning well given that whatever the national considerations, the excise duty is charged where the product is released for consumption – but to the effectiveness of the thresholds in helping the MS to set national policy objectives. As reduced rates are therefore irrelevant for producers of wine, ethyl alcohol and IP this impact assessment will not focus these products.

The key driver of the dysfunctional application of reduced rates relates to the obsolete and unclear provisions of the Directive, which is discussed in detail in Annex 11.

2.5.2. Consequences: who is affected and how?

Much of consequences of this problem, particularly with regard to the businesses and to some extent also the national administrations, has been explained in detail under the core problem definition. This is because how some of the stakeholders are affected constitutes precisely the problem at stake in the present impact assessment. To avoid repetitions, this section summarises the main impacts under headings relevant to specific type of stakeholders.

Member States authorities The reduced rates for small producers and low strength alcohol reduce the revenue MS collect from excise duties. However, all customs authorities interviewed during the Studies supporting this impact assessment considered that the reduced rate schemes did not generate large costs for the public budget. Similarly, in terms of administrative burdens for businesses and enforcement costs for public authorities, the Studies confirmed that the reduced rates did not require unnecessary efforts, by either businesses or customs. Enforcement costs with respect to domestic producers were considered to be minimal by all tax and customs authorities interviewed. The UK noted in the context of the Ramboll evaluation that reduced rates for small cider makers contributes greatly towards rural economies, has a minimal impact on government revenue and has no adverse impact on intra-EU trade.

In the cross-border context, as described under the problem definition, the increase in the number of small producers of beer, their complex business structures and the increase in cross-border trade generates some problems for enforcement and implementation, as national authorities must determine if the producer is entitled to the reduced rates.

The uncertainty in the interpretation of 'legally and economically independent' results in legal proceedings and therefore costs for stakeholders. No specific estimates or anecdotal evidence was provided by the MS, which most likely stems from the fact that the administrations or companies do not keep such readily available statistics and disaggregating from other data is difficult.

Businesses Findings show that the vast majority of active brewers, 97% in the EU, are eligible for the reduced rates scheme for small brewers and therefore subject to the legal uncertainties highlighted above. This uncertainty may hinder expansion and development of small brewers.

The competitive position of small cider makers vis a vis large producers is similar to that of small breweries. However, while small beer producers are entitled to enjoy reduced rates, the small cider makers are not. While the minimum rate for cider is zero, most MS with a traditional cider market apply a positive excise rate. As a result the zero excise rates is only applied to 9% of cider consumption.

Indeed, given the industry similarities between beer and cider, the competitiveness of the small cider makers could be explained through the proxy analysis done for the small breweries. The 21

latter, in the course of the studies supporting this initiative, were asked whether reduced rates supported their competitiveness or if the reduced rates were largely appropriated by distributors or passed on to consumers – resulting in a neutral effect on small brewers overall. Small brewers considered that the provision supported directly the competitiveness of small producers as the tax reduction was not passed through the value chain down to the consumers, and that the rebate was effective in counterbalancing lower costs enjoyed by large companies, in particular because of economies of scale and market access barrier.

While reduced rates have a clear positive impact on Small and Medium-sized Enterprises (SMEs) competitiveness, their effect on the entry rates in the beer industry is not univocal. There seems to be a trend towards the growth of the micro and small brewery segment, which is, according to businesses’ view, largely driven by market demand, and which is even across countries, regardless of whether they have implemented the reduced rates or not. In FR and the UK, where the discount for microbreweries is significant (50% of the standard rate), their number has more than doubled over the 2010-2015 period (annual growth rate of respectively 16% and 19%). In AT, the number of microbreweries remained stable (+13% over 5 years); however, the discount for microbreweries in this country is significant (40% of the standard rate). In Italy, where there are no reduced rates, the number of microbreweries almost doubled in the 2010-2015 period. While businesses consider that reduced rates support the entry of new players, these data suggest that the provision of reduced rates is neither a necessary nor a sufficient condition, and that other national factors are also at play (again, consumer demand, as well as industry structure, market stability, type of beer consumed by the population, competition from other beverages). Overall the Ramboll evaluation concluded that it is unlikely that the presence of reduced rates creates market distortions by unduly advantaging smaller firms that benefit from the rates.

The threshold for low strength beer to apply reduced rates is low and as a result there is little incentive to develop this sector. Beer producers interviewed noted that producing low strength beers cost more than producing standard beers and therefore only certain large producers can absorb this cost. Low strength beer may also taste differently from regular beer due to the ABV.

Consumers It is possible that small producers – of beer or cider - who fail to receive the reduced rates due to a different interpretation of 'legally and economically independent' or due to the lack of corresponding provisions allowing for duty reduction, may choose to absorb this cost as it has been described above. However as these are small producers, with tight margins, it could be assumed that the extra excise duty would (have to) be at least partially passed on to the final consumer. The small breweries interviewed for the supporting studies did not seem to confirm this, claiming small brewers are most likely to produce craft beer, as opposed to the mass products mostly marketed by large companies. As a consequence, price levels are different, and this reduces the incentive to pass-on the tax discount in order to remain competitive vis-à-vis larger players. Further empirical evidence is provided by an industry study on British small brewers, where most of the respondents indicated that the excise reduction was kept within the firm (e.g. for investment), and only 12% indicated that it led to a price reduction18. As above, it can be assumed that the cider market would behave similarly and is similarly impacted.

2.5.3. How will the problem evolve (baseline scenario)

With respect to small brewers, in most of MS analysed in-depth in the Studies for this initiative, the number of microbreweries is growing quickly. Even though the rate of growth is likely to

18 SIBA, “British Beer: The Report on the 2017 Members’ Survey of the Society of Independent Brewers”

diminish in the future as the market achieves a higher level of maturity, there is no indication at this stage that the phenomenon is halting19. So far, the growth in the number of small brewers was not matched by a parallel increase in their market share in the beer market. As such, there is a limited expectation that market effects (i.e. competitive distortions), costs to the public budget, or health impacts would become more prominent in the future.

At the same time, as the sector of small brewers achieve maturity and some players grow in size, it is likely that (i) cross-border trade flow augments, so that the not always smooth functioning of the scheme in MS other than that of establishment becomes a more significant problem; and (ii) the commercial relationships become more complex, and more forms of cooperation could be part of the grey areas identified above, if the provisions are unchanged.

With respect to small cider makers, the current imbalances within the markets will remain, should no change to the current situation be introduced. Although the cider industry is small and traditionally EU based20 compared to other alcoholic beverages, the industry is one of the fastest growing in some MS21. As the industry grows, it is likely that players may wish to increase their cross-border trade to remain competitive. This may be difficult in the absence of reduced rates.

In terms of the thresholds for alcoholic strengths of certain beverages, many large beer producers are currently launching new beer with alcohol strength of 3.5% vol. Although above the reduced rate threshold these new beers could be considered as a competitive product to the current low strength beers on the market. Some MS (FI, SE, DK, IE and UK) support this extension and other MS pursuing healthier drinking policies may wish to promote this, but the Directive does not favour such objectives. Some other stakeholders argue that the application of reduced rates to low-strength alcoholic beverages may increase alcohol related harm as more affordable products may eventually encourage consumers to drink more.

2.6. Problem 4 – Unclear provisions to measure the Plato degree of sweetened / flavoured beer

2.6.1. The problem and its EU dimension

Art. 3(1) of the Directive allows for levying excise duty on beer with reference either to the Plato degree or ABV of 'finished product'. The term 'finished product' is not defined in the Directive and this results in three different interpretations when it comes to measuring the Plato degree of sweetened/flavoured beer (i.e. mixture of beer with non-alcoholic additives or beverages). The addition of sugar/flavour in the beer after fermentation may artificially affect its Plato degree, as the Plato method seeks to estimate the concentration of extract in a fluid as a percentage by weight. The three different methods result in non-uniform measurement of the degree Plato; depending on which approach is chosen. This, inevitably, leads to differences in the excise duty applied to products which can have the very same alcoholic content.

Different excise duties will be mirrored in retail prices and consumption of such products. Indeed, some beer producers have reported that accounting for the added sugar when measuring the Plato degree is technically wrong, and can lead to unfair competition among them and in particular in

19 British Beer & Pub Association, “Small Brewer Relief and the impact on future market structure – Discussion paper”, 2016. 20 The European Cider & Fruit Wine Association, European Cider Trends 2017 note that 57% of consumption in 2016 was in Europe. North America and Africa account for 11% of consumption each. 21 Per the European Cider & Fruit Wine Association, European Cider Trends 2017, the 5 year compound annual growth rate (2011 – 2016) was 156% in CZ, 122% in PL and 102% in RO.

comparison to beer mixes which contain artificial sweeteners instead of sugar; for the latter products the increased excise duty would not apply.

The problem has also led to conflicts between beer producers and tax authorities, which require legal proceedings and rulings and entail litigation costs. In Germany, a brewer went to court in 1997 seeking to have its radler (a type of sweetened beer) taxed based on the ‘real extract’ rather than the ‘present extract’ (see Annex 13 for further information on the approaches to measuring the Plato degree of sweetened/flavoured beer). The national court finally decided against the brewer’s pleads only in 2004. Recently, a similar case has been brought to court by a Polish brewer (see Box 5). This case was referred to the CJEU (C-30/17 - Kompania Piwowarska) but only following 12 years of local legal proceedings in PL regarding the way in which excise duties on such beer should be determined.

Box 5 - Calculation of excise duty on sweetened/flavoured beer: The Polish case (C-30/17 - Kompania Piwowarska)

In this case the national Polish court requested a preliminary ruling concerning the calculation of excise duty on sweetened/flavoured beer. A Polish beer company producing sweetened/flavoured beer disagrees with the Polish tax authority on the measurement method. The different views of the brewer and the Polish tax authority can be summarised as follows.

The brewer argues that the strength of the sweetened/flavoured beer in Plato degree should be measured accounting for the ‘real extract’ (method B1) rather than ‘present extract’ (method B2) of the finished product. Including the sugar added after fermentation in the extract figure would be a technically wrong measurement, because this sugar does not add to alcohol formation. By contrast, the Polish tax authority requires method B2, i.e. measuring the Plato degree on the basis of the present extract, including the sugar added after fermentation.

The Polish case clearly demonstrates the importance of the problem for both beer producers and tax authorities. By adopting the brewer’s approach, the beer producer (tax authorities) must pay (receive) PLN 87.8, whereas by adopting the tax authority’s approach it must pay (receive) PLN 109.8 (figures correspond to the example provided in Annex 14) per hectolitre of beer. Different interpretation of the way of applying the Plato method to sweetened/flavoured beer can lead to differences in excise duties for the same product.

This example demonstrates the legal uncertainty for businesses associated with the co-existence of the different measurement methods, which constitutes an additional aspect of the problems related to the measurement of Plato degree of sweetened beer. The key driver of this problem is the divergent interpretations of the term 'finished product', which is discussed in Annex 12.

2.6.2. Consequences: who is affected and how?

Member States authorities MS collect revenues from the excise duties. As shown in Annex 13 and 14, method B2 yields the highest excise revenue for the authorities. Moreover, approach A and B1 generate some problems for enforcement, as national authorities are reportedly unable to measure the Plato degree of the base beer, the real extract or the present extract by analysing the content of the bottled beer; any such checks must be done at the brewery by measuring both the Plato degree of the base beer and the quantity of base beer included in the end-product. In this context, enforcement problems become more prominent when it comes to applying excise duty on sweetened/flavoured beer moved from another MS, as tax authorities could hardly perform checks in breweries based in a different country. Approach B2 was therefore found to be the only one allowing for proper checks by customs laboratories, thus reducing room for tax fraud. It is also the method applied in the majority of the sample MS even though the industry is of the opinion that it is technically incorrect.

Finally, as described above, the legal uncertainty and the differences in interpretations lead to legal proceedings and therefore costs for tax authorities and beer producers. 24

Businesses Some beer producers claim that the discrepancies may ultimately lead to distortions of competition caused by artificially – or mathematically - changing the Plato degree without altering the alcoholic content. There is no market data available which would distinguish the different types of sweetened/flavoured beers while taking into account their methods of production and measurement of the alcoholic strength to confirm those claims. The businesses interviewed in the context of the two Studies regarding the revision of the Directive often did not have readily available or shareable market analysis. A quick calculation of the different excises theoretically applied to sweetened/flavoured beer with alcoholic strength measured using the different methods, nevertheless highlights the disparities of treatment and potential for distortion (see Annex 14 for details).

On the other hand, it should be noted that sweetened/flavoured beer producers are free to use sweeteners (e.g. aspartame) to sweeten their products instead of sugar, if they want to avoid extra taxation on the added sugar. Unlike sugar, sweeteners do not increase the Plato degree when switching from approach B1 to approach B2. Tax authorities argue that given that only a few brewers use sweeteners instead of sugar, it shows that the extra excise duty is not a high burden for them. By contrast, brewers explained that the choice to use sugar rather than artificial sweetener is driven by marketing considerations, e.g. using only natural ingredients, rather than by cost considerations, e.g. tax savings.

Consumers It could be argued that, in order to keep the competitive edge, the producers of sweetened/flavoured beer may have to choose to absorb the extra cost. Assuming nevertheless the excise duty is consistently passed-on to consumers in the retail price, it would affect the competitiveness of products and the related demand, causing ultimately potential distortion of the market. In any case, the additional cost of higher excise duty would not disappear and would have to be borne by one or the other party.

Any change in excise duty reflected in a change in price is expected to impact the consumption of sweetened/flavoured beer. This in turn can, albeit to a minor extent, engender public health policy issues. For instance, sweetened/flavoured beer is thought to be more attractive for women and young consumers, neither established beer drinking groups, which is confirmed by the fact that 40% of radler drinkers are new to the beer category. However, research has shown that marketing plays a greater role in attracting these consumer groups than the actual content/taste of a beer22. On the other hand, radler contains less alcohol (2-2.5% vol) than standard beer, so it may be desirable to promote a shift towards beverages containing less alcohol. This would eventually reduce the overall alcohol intake and ultimately result in positive public health impacts.

2.6.3. How will the problem evolve in case of no-EU action (baseline)?

Even though the EU market for sweetened/flavoured beer is relatively small – around 2.7% of the overall beer market in 2015 – it is growing faster than the beer market itself, which has actually stagnated in many industrial economies.

Sweetened beer is part of a strategy of brewers to innovate and regain market share; even the mainstream beer brands like Heineken or Peroni have introduced sweetened/flavoured beers, and

22 WHO ‘Global status report: alcohol and young people’, 2001; and: ‘Beer and Health: Moderate consumption as part of a healthy lifestyle’, at http://beerandhealth.eu/wp-content/uploads/2016/07/beer-and-health-web.pdf (last accessed on 10 July 2017). Stakeholders have also confirmed this statement.

especially radlers, in recent years. The IWSR database reports a market growth by 6% between 2015 and 2016 for sweetened/flavoured beer, and projects a market growth by 8.5% in 2017. A study23 forecasts steady growth for sweetened/flavoured beer in Europe to 2020. Therefore what may seem a problem of a limited, local scope today, unaddressed could evolve to a much bigger impact on the future functioning of the internal market, even though estimates are not available.

The CJEU is called to rule on whether the Plato degree of sweetened / flavoured beer should be measured by considering the 'real extract' (approach B1) or the 'present extract' (approach B2). The awaited judgement of the CJEU on the prejudicial question of the Polish court may contribute to addressing - and eventually clarifying - the policy problems. At the moment, the baseline scenarios will be different for the MS, depending on which method of measuring Plato degrees they apply. Regardless of the ruling, some MS will need to adapt the methods in order to comply with the ruling. The extent to which the CJEU ruling will change the status quo is therefore presently unknown.

2.7. Conclusion

It is apparent from the analysis of the problems above that the functioning of the current system for alcohol and alcoholic beverages is causing disturbance to both MS and businesses. These are problems that are exacerbated by the increase in cross-border activity that is the result of globalisation of the economy and the extension of the EU (from 12 to 28 MS) since the Directive was adopted. In some cases this also provides greater opportunities for fraudsters.

3. WHY SHOULD THE EU ACT?

In analysing the problems and the problem drivers it is clear that the Directive in general works well and provides an EU-wide system of uniformity and harmonised conditions that are necessary to ensure the proper functioning of the internal market. Despite the shortcomings described no alternative national, bilateral or other international initiative would provide the same level of effectiveness in terms of the functioning of the internal market and the monitoring and control of excisable alcohol, and significant added value consequently accrues from establishing common definitions and rules of alcohol and alcoholic beverages for excise purposes at EU level.

When looking at the provisions related to denaturing alcohol and in particular PDA, the source of the current complications lays precisely in the absence of clear rules at EU level. Aligned to this, and because of that ambiguity, the MS are interpreting those current rules differently, and businesses therefore take advantage of the more flexible approaches used in certain MS. There is a lack of clear understanding of the rules on mutual recognition of denaturing methods between MS, which also causes administration problems for authorities and businesses alike. MS themselves highlighted the need for clear rules on the exemption of denatured alcohol. One MS noted for example that the “definition of rules at this [EU] level is of utmost necessity, otherwise each MS will have its own system, according to its national interests, and that will only complicate matters.” Another MS remarked that “a common system established at EU-level will help the functioning of the common market and facilitate equal treatment. However, any rules must be detailed and clear enough to ensure they are interpreted the same way in all MS.” The evidence from both Studies showed that clear rules, common for all MS would protect the single market. No bilateral or multilateral agreements could have the broad EU impact.

23 http://beer.drinks-business-review.com/news/demand-for-low-or-non-alcoholic-beer-to-grow-in-europe-through-2020- 030117-5708450 (last accessed on 10 July 2017)

Decisions taken unilaterally by MS, such as issued BTIs for certain alcoholic beverages, create additional complexity. A solution that would clarify the scope of the current categories in agreement of all MS would provide a much more effective solution. Although rulings of the CJEU established criteria to classify borderline products from genuine OFBs, the subjectivity of the criteria has magnified the classification uncertainties. The Ramboll evaluation remarked that although little quantifiable data was available for analysis, taking up effective measures to resolve difficulties in classifying alcoholic beverages for excise purposes would reduce administrative costs both for the Member States’ administrations and for the economic operators involved. It concluded there is significant added value from establishing common definitions of alcohol and alcoholic beverages for excise purposes at EU level.

When it comes to the reduced rates for small breweries, used by many MS, the lack of clarity of the term 'independent brewer' and the cross-border implementation of the reduced rates is problematic. In the Ramboll evaluation twenty MS strongly agreed that setting the basic rules at EU level would support the application of an uniform approach and would avoid distortion of competition. Furthermore the fact that this relief does not apply to small producers of other products also distorts competition within and between MS.

The reduced rates for low strength alcohol are irrelevant for most beverages as a result of other Union law. The threshold for beer does not encourage brewers developing low strength beers. BE and SE supported reduced rates for low strength alcohol in the Ramboll evaluation as they allow for the promotion of alternatives containing less alcohol. In their opinion this is better for consumers' health and is working towards a system of taxing products based solely on their alcoholic content.

With regards the measurement of Plato degree of sweetened / flavoured beer, the source of the current complications lays precisely in the absence of clear rules for 'finished product' at EU level. Because of that ambiguity, the MS are interpreting those current rules differently.

As with the subsidiarity test, it is not possible for MS to address the problems and problem drivers in isolation without a proposal to amend the structures Directive.

In conclusion, if the problems at hand are to be addressed in a coherent and meaningful fashion it can only be achieved through a legislative proposal supported by some non-legislative guidelines. Therefore, it is necessary for the Commission, which has responsibility for ensuring the smooth functioning of the internal market and promoting the general interest of the European Union, to propose action to improve the situation. The legal basis is Art. 113 of the Treaty on the Functioning of the European Union (TFEU).

4. WHAT SHOULD BE ACHIEVED?

As explained in detail under the problem definition, given the broad scope of Directive 92/83/EEC covering a variety of products and provisions, the problem areas under this initiative are, for the main, very divergent from one another, requiring dedicated specific analysis. The complex structure of this report is illustrated in Figure 2. As a result, the objectives are also drawn up in such a way that they correspond only to specific problems/drivers.

4.1. General objectives

The spirit of Directive 92/83/EEC and its general objective is the proper functioning of the internal market for alcohol and alcoholic beverages. In the context of this initiative, this objective is complemented by two other general objectives, which were identified applicable during the evaluation: safeguarding the revenues of the MS and contributing to protection of human health. 27

The last two objectives, although not directly relevant to all problem areas, are particularly important for some of them, as shown in Figure 2. It was therefore important to have them included in the scope of the analysis and propose measure with the aim of achieving them.

The general objectives behind the initiative are therefore as follows:

 ensuring the proper functioning of the internal market for alcohol and alcoholic beverages, free and undistorted movement of such goods within the EU (Art. 26 and 113 TFEU);  safeguarding the revenue of MS;  ensuring human health protection in Union policies and activities (Art. 168 TFEU).

4.2. Specific objectives

The general objectives translate – albeit not one-to-one (see Figure 2), into the specific objectives, which can be defined as follows:

 ensuring fair treatment and similar economic conditions for businesses across all alcohol sectors, including small producers of all alcohol types;  preventing and correcting any distortions of competition in the application of the exemption for different types of denatured alcohol, of the excise duty for sweetened beer, and of the reduced rates for low strength alcohol and small producers;  providing clear rules on the scope, classification and calculation of excise duties for businesses and MS  providing clear and efficient conditions to determine denaturation procedures for all types of denatured alcohol;  reducing administrative burden and compliance costs for businesses and tax authorities, and providing legal certainty specifically in the area of classification and the exemption for denatured alcohol;  strengthening the fight against fraud and tax evasion (including excise duty circumvention), through clear and consistent framework governing the calculation and collection of excise duties.

5. WHAT ARE THE VARIOUS OPTIONS TO ACHIEVE THE OBJECTIVES?

5.1. Link between problems/drivers and options

As detailed in section 2.1, the problems analysed in this report touch upon 4 distinctive areas: (i) exemptions for denatured alcohol, (ii) classification of certain alcoholic beverages, (iii) reduced rates for small producers and low strength alcoholic beverages, and (iv) measurement of Plato degree of sweetened/flavoured beer. These distinct problems, and their underlying drivers, need to be addressed in different ways, which influences the chosen aggregation of impacts into individual sets of measures targeting specific issues to be bundled together at the end of the analysis into packages.

To illustrate this better, improving the provisions of Directive 92/83/EEC may be the solution to resolve some of the problems; for others, the solution may be found in legislation that is outside the scope of the present initiative. There are also specific areas where no alternatives other than acting/no-acting could have been identified. The report considers all possible policy options but focuses its analysis on the ones, which have been retained for the policy-makers. The reasons for discarding some options early on as well as considerations and constraints behind others are presented under each cluster.

For better illustration of the problems, their drivers, objectives and corresponding options are presented in Figure 2 below. The baseline scenarios have not been included in this figure although they are systematically described under each policy cluster and constitute the framework against which all options will be assessed.

Figure 2 – Overview of the intervention logic

5.2. Dysfunctions in the application of exemptions for denatured alcohol

When looking to resolve the issues with denatured alcohol, there is a need to balance between harmonising the understanding of the provisions to reduce the effects of the differing interpretation, maintaining flexibility for producers and users of denatured alcohol to have denaturants that match their products, and ensuring that customs authorities can implement sufficient control to limit the risk for abuse of the exemptions.

Ideally full harmonisation of CDA formulations would be the obvious policy option to resolve the legal uncertainties that persist around the mutual recognition of CDA. This would entail:

 Agreement on a single formulation, containing the same denaturants in the same concentration for CDA across the entire EU;  Elimination of all remaining national formulations;  Potentially a significant change in the wording of Article 27(1)(a) and 3 and 4, to reflect a new procedure for defining the common formulation, which would supersede the current process of notification by the MS.

There is strong opposition from a limited number of MS to the full harmonisation of CDA formulations. Even those in favour of full harmonisation may wish to retain control over possible future changes and therefore would not agree to a change to the notification process of Articles 27(3) and (4). Furthermore, findings of the Ramboll evaluation do not suggest that there should only be one denaturing method, neither to prevent fraud, nor to ensure fair competition between economic operators. Therefore this option will not be assessed further in this impact assessment due to the fact it is unlikely to be feasible at this time.

5.2.1. Option 0 – baseline scenario

The adoption of Regulation 2017/2236 on the mutual recognition of procedures for the complete denaturing of alcohol for the purposes of exemption from excise duty will greatly reduce problems arising from the unclear rules on recognition of CDA formulations. The possibilities for fraudsters to use the 'weakest' formulation will be reduced by the replacement of national formulation by the Eurodenaturant and thereby reducing the risk of fraud with CDA overall. However, the problems will not be fully eliminated and MS could re-introduce national CDA formulations, if they wish to.

The proliferation of national approaches to PDA will continue and possibly intensify for biofuels, which accounts for the largest proportion of PDA. Divergent interpretations in the area of PDA are likely to remain despite the exploratory work carried out by the Fiscalis Project Group and the uncertainty for cross-border trade will continue.

5.2.2. Option 1 –clarify mutual recognition of CDA

This option would clarify the rules in the Directive for mutual recognition of CDA in order to eliminate divergent interpretations. The identified possible approaches to clarify mutual recognition are the following:

'Hybrid' mutual recognition (option 1.a): Each MS would have to recognise CDA produced in another MS using the formulations notified by that particular MS, but not those notified by any other MS. This would mean that MS retain control over the CDA produced within their territories, while being obliged to also exempt any CDA legally produced in another MS.

Full mutual recognition (option 1.b): All MS would have to recognise all procedures notified by all MS, irrespectively of where the alcohol was produced / denatured. This would effectively

eliminate all national differences, and mean that a formulation notified by a given MS could be used by producers across the EU, and the resulting alcohol recognised as completely denatured by all MS.

Limited mutual recognition (option 1.c): Each MS would only be obliged to recognise its own formulation(s), irrespectively of where the alcohol was produced / denatured. This would mean that a producer in a given MS would have to use different CDA formulations for different national markets.

To illustrate the difference between the three approaches, consider the example of the remaining CZ national formulations: under the most ambitious approach 1.b, all MS would have to allow their economic operators to use these formulations. Under the approach 1.c, alcohol denatured in CZ using these formulations would not have to be recognised as CDA by any other MS, although producers in other MS would be able to produce and export this to CZ as CDA. Under the approach 1.a, the CZ formulations could only be used in CZ, but alcohol denatured in CZ using these formulations would have to be treated as CDA and therefore exempted by all MS.

The approach 1.b would effectively turn the remaining national formulations into additional Eurodenaturants, which many MS would not accept due to the concerns over the robustness of some formulations, which in their eyes hampers the national objectives of combatting fraud or protecting health. Approach 1.c on the other hand would be more restrictive than the current situation and authorities would face enforcement difficulties. Due to the lack of political feasibility for full mutual recognition and the restrictive characters of the limited mutual recognition, both of these options are discarded and will not be analysed further. The remaining option 1.a will be hence on presented simply as Option 1.

5.2.3. Option 2 – Harmonisation of PDA formulations

While full harmonisation is the preferred policy option to resolve the problems for PDA, this is currently not feasible despite the exploratory work carried out with the Fiscalis Project Group. This is due to the numerous national approaches which are currently extremely different and MS have indicated that they are not prepared to substantially alter their approach. Therefore this option will focus on partial harmonisation of PDA formulations and would consist of developing an harmonised list by the existing FPG or another expert group, that is applicable across the EU. This would enable MS, subject to certain conditions, to authorise different formulations, not included on the list, for specific uses where the fiscal risk is demonstrably low. This option would involve both a regulatory and non-regulatory aspect. The new approach would be included in the Directive. In addition criteria, guidelines and procedures would need to be adopted for determining low fiscal risk and amending the harmonised list. For this reasons the non-regulatory measures alone would not be a viable option to be deployed individually and is not considered as such in this analysis.

5.2.4. Option 3 – Confidence / capacity building measures

This option focuses on increasing trust and confidence between MS. Some stakeholders believe the current difficulties regarding the treatment of PDA are created due to a lack of trust between MS authorities. This arises due to the different supervisory approaches and a suspicion that some countries' procedures and formulations are ineffective. It has been suggested that this could be resolved by increased information sharing, working visits, twinning or exchanges. A separate option proposed by the Study involves the creation of a national PDA database. This option will not be analysed as the European alcohol denaturant database which is accessible to MS national authorities and the Commission, already exists. Currently this database, which had fallen into disuse by some MS, is being updated by all MS. An update to the database would

enhance transparency and allow economic operators to check whether a given formulation they would like to supply or procure is authorised in the relevant MS, thereby enhancing legal certainty and reducing barriers to trade.

5.2.5. Option 4 – Legal clarification of terms relating to PDA

The purpose of this option to clarify the legal base that relates to PDA (Art. 27(1)(b)). Overall the clarity of the legal base could be improved. In addition the terms 'used for the manufacture of' and 'finished product' would be clearly drafted. This would reduce the risk of divergent / arbitrary interpretations across the EU and ensure equal treatment of goods containing PDA.

The clarification of 'finished product' is particularly challenging as a finished product across the various product groups (i.e. cosmetic product and screenwash) is extremely diverse. The clarification would make reference to a 'recognisable finished product' or 'finished product' in order to provide MS with flexibility for the various product groups using PDA. This option could also define a quantitative line above which a product containing denatured alcohol must always be classified as CN 2207 20 00 (and therefore be considered an excise good and treated as such, similar to the clarification of mixtures containing ethyl alcohol used as raw material to produce fuels for motor vehicles24.) This could be included as an amendment to the Directive or defined via a Commission Implementing Regulation (CIR) and / or a note to the CN). This option could also require such alcohol to move in accordance with Chapter IV of Directive 2008/118/EC. This option will focus on the latter amendment to the Directive in the interest of clarity and legal certainty.

All options put forward and retained are complementary and could be deployed together, affecting different aspects of the problem with denatured alcohol.

5.3. Dysfunctions in the classification of certain alcoholic beverages

5.3.1. Option 0 – baseline scenario

It is expected that national custom authorities will continue to adopt alternative methods for classification to deal with the subjective criteria given by the CJEU. It could also be envisaged that to solve the dilemmas created by innovative products which it is generally agreed should not benefit from the preferential treatment, MS could resort to unilaterally changing the rate of excise tax of OFB in order to bring the expected tax due under this category approximately into line with that applying to beverages of similar strength and falling under ethyl alcohol. If it came to this, MS acting purely to protect their national interests, would further erode the very rationale for the establishment of the category. As the current specifications of the EMCS lack the OFB category to distinguish OFB from wine (W200), moving away from an equivalence of taxation between wine and OFB would create an inconsistency within the system. These approaches are non-harmonised and the risk of different legal interpretations is likely to persist or grow, leading to different classifications and more abuse.

The adoption of a new note to Chapter 22 of the CN code to guide the classification may assist in reducing the uncertainty. However, the CN code is outside the remit of excise duty authorities

24 Commission Implementing Regulation (EU) No 211/2012 of 12 March 2012 concerning the classification of certain goods in the Combined Nomenclature, OJ L 73, 13.3.2012, p. 1–2) and 626/2014 (CIR (EU) No 626/2014 of 10 June 2014 amending Annex I to Council Regulation (EEC) No 2658/87 on the tariff and statistical nomenclature and on the Common Customs Tariff, OJ L 174, 13.6.2014, p. 26–27).

and furthermore the notes to the CN code are not binding. Therefore this could furthermore increase the risk of disparities of interpretation.

Monitoring and evaluation of the French Soprano classification system launched in 2017 (see section 2.4.3) is necessary before proposing an EU wide adoption. Besides the fact, Soprano is only in its infancy, this platform is based on the FR approach to classification and therefore the risk of different legal interpretations together with ensuing disputes and incentives to develop products exploiting the ambiguity, persists.

Competitive advantages for businesses with a favourable tax classification obtained by 'classification shopping' are likely to continue. Moreover, due to the high costs, businesses will remain risk-wary towards the placement of new products on the market without formal classification from customs authorities.

MS will also continue to adopt national legal and administrative provisions to ensure a certain margin of tolerance for the addition of AFC. It is expected that MS with no formalised approach will also adopt domestic measures for AFC, with increasing cross-country disparities. The ambiguity with the legal text of the Directive would persist.

5.3.2. Option 1 – clarify the excise duty structure for 'borderline' products

This approach consists of refining the current definition of certain excise duty categories so as to reduce the risk of disparities of treatment and/or unduly favourable treatment of ‘borderline’ products, but without changing the five-category fundamental structure of the Directive.

The tax classification of these products would not be so strictly determined by the customs classification. The excise definition of products should evidently remain linked to the CN heading, but the criteria that today determine if a borderline product should fall under Art. 20 or not could be established explicitly in the tax legislation rather than derived from the prior CN code. Under the current system it is the customs classification which determines the excise duty category. Once a beverage is classified as CN 2208 (undenatured ethyl alcohol) it can be taxed only under Art. 20 (ethyl alcohol). If classified as CN 2206 (OFB) it may fall under Art. 12 (OFB) or Art. 17 (IP) depending on its strength, but not under Art. 20.

This would translate into introducing in the Directive the same CJEU principle that currently inform CN classification, which establishes that a fermented-base beverage that has lost its essential character (taste, smell, and appearance) can be assimilated to a distilled-base beverage, and subject to excise duty in accordance with Art. 20. This approach would require an amendment to the text of the Directive, so that:

 products that have lost their essential fermented character would be excluded from the scope of Art. 12 and 17; and  products classified under CN 2206 of any ABV strength would be allowed under Art. 20 (the denomination of the category might be revised accordingly).

Under this approach, MS may consistently tax any ‘borderline’ product under Art. 20 that is considered as having lost its essential fermented character, regardless of the fact that it comes under CN 2206, with or without a BTI.

A further clarification of the excise structure would propose adopting a flexible approach toward AFC, allowing the addition of ethyl alcohol of agricultural origin to products of ‘entirely fermented origin’ (wine and OFB) to dilute or dissolve colorants, flavourings or any other authorised additives and not exceeding the dose strictly necessary. The principle can be established in the Directive in generic terms, as in Regulation 251/2014, or setting an upper limit

to the maximum contribution of AFC to the total ABV of the final products. This clarification would have limited impact on the disparities of treatment of 'borderline' products and may have unintended consequences for certain aromatised wine products. This element will not be assessed further in this impact assessment.

5.3.3. Option 2 – introduce a differentiation in the OFB tax category

This policy option consists of a possible extension of national approaches to the EU-level, which are  distinguish for tax purposes traditional cider and other products defined in country-level sectoral legislation, from all other generic OFB, including ‘mass-market’ cider and the like  apply additional consumption taxes on specific categories of mixed drinks to deter their consumption

These approaches aim to differentiate the OFB products that arguably correspond to the original definition and intention of the legislator from the ‘novel’ products that have been opportunistically designed to fit into it or simply that do not fit elsewhere. In fact, the two existing approaches have the same objective and result. The only difference between them regards which ‘sub-category’ is separately defined and excerpted from the standard one – i.e. the ‘mixed drinks’ (intended as ‘pre-mixes, alcopops etc.) or the ‘cider and perry’. In visual terms, the two approaches can be represented as in Figure 3 below, where their difference concerns where the demarcation line is drawn, namely:

 Line A (Option 2.a): cider and perry (and specific OFB like mead, hydromel, certain fruit- wine etc.) v. Other OFB (including mixed drinks and possibly certain ’borderline’ cider drinks).

 Line B (Option 2.b): mixed drinks (pre-mixes, alcopops and the like) versus cider, perry and any other non-mixed OFB of any kind (‘traditional’ or not).

Figure 3 – The two possible approaches for differentiating the OFB category

Cider, perry etc. Other OFB Mixed drinks Premixes, alcopops etc.

A B The demarcation Line A would require adoption at EU level of a harmonised definition of cider, perry and the other OFB that correspond to the original scope of this category, matching as much as possible with the existing national definitions for these products. The demarcation Line B would require a harmonised definition to be adopted at EU level defining a mixed drink and the relevant criteria to allow for such a categorisation.

With the exception of FR, where both differentiation lines are in place, all other MS have opted for only one distinction. In this impact assessment, the third approach (based on the French practice), is not proposed as introducing two differentiations would excessively fragment a category that is currently small.

5.3.4. Option 3 – provide operational definitions, criteria and methods

Common rules and criteria would be necessary to establish/determine when a product has actually lost its essential fermented character irrespective of any legal changes to the excise

classification. Such criteria should not be in the text of the Directive but defined in detailed operational terms in guidelines, recommendations and/or explanatory notes to CN nomenclature.

Many tax administrations interviewed in the context of the supporting studies were of the opinion that a proper operationalisation of these criteria or any solution sought at the level of excise duty classification would fail, because there are uncertainties in the primary underlying CN classification. What is currently included in the explanatory note to CN 2206 00 reportedly leaves a wide margin for subjective interpretation. Simply introducing the CJEU jurisprudence principles in the Directive would still require clear, agreed, and robust criteria and analytical methods to be in place. Such criteria, conditions and methods could be established: at the level of CN explanatory notes (option 3.a) or, in any case, within the customs classification system (revision of the CNEN 2206 00) where a robust distinction between fermented alcoholic products that may fall under CN 2206 and those that should be considered CN 2208 could be provided. A customs expert group (Customs 2020 Project Group) is currently discussing and drafting an implementing regulation to create a new additional note to Chapter 22 of the CN code to guide the classification of these alcoholic products. This note will focus on distinguishing between the CN codes and will also touch upon classification of new products using cleaned-up alcohol. The draft implementing regulation is scheduled for vote in the Committee meeting of June 2018. through non-binding guidelines (option 3.b) –guidelines would be developed by a joint technical working group and adopted at ITEG level. These guidelines should, among other things:

o establish the criteria to differentiate between a ‘genuine’ fermented beverage and a beverage that has lost its essential character, which should be classified otherwise and provide guidelines to indicate how to weigh and balance the different aspects; o set a threshold for the amount of distilled alcohol that can be added to a fermented beverage both in terms of contribution to the total ABV and/or overall volume of the end- product, and other parameters related to the appearance and taste of the product; o establish if, and to what extent, the addition of other substances like water, sugar, cream etc. may per se affect the fermented character of a beverage or not, and the criteria thereof; o establish analytical parameters to deal with ‘cleaned-up’ alcohol, both as an end-product or a base for other beverages; o define common analytical methods to assess the composition of products in order to improve detection capacity and reduce uncertainties in laboratories’ outcome. These measures presented above do not require a revision of the Directive and can be self- standing options. They are however not strictly alternative to options 1 and 2, but rather complementary and in some case a pre-requisite for a successful implementation of the proposed Directive amendments.

5.3.5. Option 4 – amend other EU legislation

Sectoral regulation for cider and other specific OFB (option 4.a). This option envisages adopting at EU-level a harmonised definition of cider, perry and other specific OFB to distinguish them from other generic OFB like mixed drink, which are arguably taking advantages of the blurred boundaries of the current excise duty definition. This would complement option 2 above, which proposes a differentiation in the OFB category and would ensure the smooth operation of reduced rates for small cider makers, if reduced rates for small producers was extended to include small cider makers (see section 5.4.3 below).

Enhance monitoring and control (option 4.b). This option proposes introducing separate codes for OFB. This would address the lack of a specific EPC for OFB which is currently merged with wine. This amendment concerns Annex II, Table 11 (Excise Product) of Commission Regulation 684/200925, as well as of the EMCS and related systems, including MS authorities and businesses' excise systems.

A further aspect of this option proposes introducing, for statistical purposes, a collection of more granular data on excise goods volumes than current data, which is articulated only on EPC, and does not cover zero-rate products. This would assist tax authorities, who currently have a limited market intelligence of novel 'borderline' products to address problems effectively and consistently. This aspect will be discarded as the current procedures and administrative arrangements in MS vary substantially and an one-size-fits-all approach is not possible. Further consultation with MS would be necessary to introduce this.

5.4. Dysfunctional application of reduced rates

5.4.1. Option 0 – baseline scenario

The unequal treatment of producers of alcoholic products other than beer and spirits will persist. MS will be unable to correct this. Divergent interpretations in the area of economic independence and the uncertainty for cross-border trade are likely to increase as the number of small brewers continue to grow, which currently shows no sign of slowing. Conflicts between businesses and authorities will persist and may even increase as business structures increase in complexity.

The application of reduced rates to low strength alcoholic beverages will continue to apply to a limited number of beer products. MS will be prevented from achieving national policy objectives of encouraging consumers away from high strength alcoholic beverages.

5.4.2. Option 1 – Increase legal certainty for small breweries

The main regulatory failures for small brewers concern (i) the existence of grey areas in the definition of economic independence; and (ii) the implementation of the provision to cross- border businesses. This option would clarify the term 'legally and economically independent' and would provide a common EU method for proving the status of producers.

Option 1.a – Normalising the definition of economic and legal independence at the EU level

To address the problems described earlier, the term 'economic and legal independence' should be defined at EU-level. Such definition would encompass the general norms and principles as well as detailed technical specification outlining the legal conditions which could determine if companies are independent or not. Some aspects have already been clarified and several CJEU jurisprudence provide for the necessary guidance, which has been developed and consolidated over the years. Any further action would therefore refer to the existing acquis as much as possible while any gaps – e.g. with regard to the forms of cooperation – would need to be addressed. This could be done by consolidating the current practices on beer brewed under license – and the present national practices – as well as contract brewing.

25 Commission Regulation (EC) No 684/2009 of 24 July 2009 implementing Council Directive 2008/118/EC as regards the computerised procedures for the movement of excise goods under suspension of excise duty.

Definition of economic and legal independence of small breweries could be done either through a legislative revision (option 1.a.1.) of the Directive or by means of a soft law instrument (option 1.a.2), such as non-binding guideline. Whereas these 2 instruments would in essence yield the same framework, they would differ in the effectiveness. It is therefore important for the analysis to retain that distinction for further comparison and identification of the preferred choice.

Option 1.b – Creating conditions for recognition of small brewers across borders

With respect to the means for proving the status of small brewers and the modalities for the exchange of information between tax or customs authorities, these could be specified along different, possibly complementary, lines:

Ex-ante approach (1.b.1): all small brewers would be identified through a uniform certificate, defined via a Commission Implementing Regulation, which would need to be presented when claiming reduced rates in a MS other than that of establishment. Such a certificate would state: (i) the brewery output level, as already communicated or available to the customs authority under tax warehouse obligations; and (ii) whether the brewer fulfils the criteria for economic and legal independence, based on additional documentation submitted by the economic operator. This certificate should be provided, upon request, by all customs authorities to all businesses up to 200,000 hl, regardless of whether they can access reduced rates in their country of establishment. This certificate could be developed through the Fiscalis programme.

Ex-post approach (1.b.2): as in the current framework, a verification of whether a non-domestic brewer meets the conditions for enjoying reduced rates would be done upon request of the authority of the MS of destination for specific players. However, these ex-post checks would be managed by an IT platform for the exchange of information, so that the authorities in the country of destination could inquire about an operator’s annual output and independence. Alternatively, each customs authority could prepare a list of breweries which are both independent and with an output below 200,000 hl. Experience with the European alcohol denaturant database shows that this option would be of limited benefit, as MS often fail to update the data regularly. This option will not be assessed further in this impact assessment.

5.4.3. Option 2 – Extending the reduced rates to small cider makers

To address the unfair competition between small producers of alcoholic beverages, this option would amend the Directive to extend the reduced rates to small cider makers.

As for the small brewers reduced rates, this reduced rate would remain optional for MS. It would be based on the definition of an independent producer and a maximum discount rate compared to the standard rate would be fixed. The maximum yearly output threshold would be set in the Directive. One possible output threshold (100 hectolitres per year) would cover micro cider makers only. The second option would apply an output threshold of 15 000 hectolitres per year, which would extend the relief to small cider makers.

5.4.4. Option 3– Revised thresholds for low strength alcohol

This option aims to amend Art. 5(1) of the Directive and allow MS to apply reduced rates to beer with an ABV not exceeding 3.5% vol (instead of 2.8% vol).

5.5. Measurement of Plato degree for sweetened / flavoured beer

5.5.1. Option 0 – baseline scenario

Under this option MS will continue to have freedom in the interpretation of the term 'finished product' when measuring the degree Plato of sweetened/flavoured beer.

A case has been referred to the CJEU (C-30/17 - see Box 5) regarding the way in which excise duties on sweetened / flavoured beer should be determined. The precise scope and extent to which the CJEU will clarify the outstanding uncertainties of the Plato situation is unknown. If the CJEU rules contrary to the existing practice of measuring Plato degree after the addition of sugar, several Member States would be required to change their approach.

5.5.2. Option 1 – Legal clarification of term 'finished product'

This option implies clarification/definition of the notion of 'finished product' and when the measurement of Plato degree should occur when it comes to beer in the legal base (Art. 3(1) of the Directive). Defining 'finished product' could be done following any of the methods currently applicable to measuring the Plato degree:

 Option 1.a – regulatory amendment of the term ‘finished product’ where it would refer to the base beer before adding any additives, i.e. Approach A of measuring degrees Plato;  Option 1.b – regulatory amendment of the term ‘finished product’ where the term would refer to the end-product that is released for consumption. This can be further subdivided in line with the two approaches B1 and B2 depending on whether the sugar/flavour added after fermentation would contribute (option 1.b.2) or not (option 1.b.1) to the Plato degree.

5.5.3. Option 2 – Guidance/recommendation on the most appropriate method to measure Plato degree of sweetened/flavoured beer

The non-regulatory option consists of providing guidance on the most appropriate approach to measure the Plato degree of sweetened/flavoured beer via non-binding guidelines/recommendation of the Commission. This option can be either alternative or complementary to option 1, in the sense that guidelines could also support the implementation of the revised regulatory provision, suggesting technical solutions, procedures and other best practices to national authorities. Similar to regulatory Option 1 guidelines/recommendations could be made based on any of the three methods currently applicable, leading respectively to sub-options 2.a, 2.a.1 and 2.b.2.

5.5.4. Option 3 – Abolish the Plato method for measurement of alcoholic strength in beer

This option would amend the Directive, so that only ABV would be allowed by MS to measure the alcoholic strength of beer.

This option would reduce the additional administrative costs that producers measuring the strength of beer using the Plato method face when they sell cross-border as they are required to report data to EMCS using the ABV method, even when the movement of goods occurs between

two MS using the Plato method. Furthermore in order to comply with food labelling requirements26, all producers must display the ABV strength on beer labels.

While this option would reduce the legal uncertainty, distortion of competition and regulatory costs, the abolishment of the Plato method would be vigorously opposed by both the industry and many MS on grounds of tradition. In fact, all relevant stakeholders interviewed for the both Studies have confirmed that there are no negative consequences for beer producers, because regulatory costs are negligible and do not constitute an obstacle in practice when it comes to selling in another MS. Taking account of the above, this option was discarded.

6. WHAT ARE THE IMPACTS OF THE DIFFERENT POLICY OPTIONS AND WHO WILL BE AFFECTED?

The impacts considered for the policy options belong to four main categories and span various categories (or even sub-categories) of stakeholders: (i) market effects (including Single Market functioning, distortion of competition, and SME competitiveness effects); (ii) regulatory costs and cost savings (including substantive compliance costs, administrative costs and enforcement costs); (iii) tax revenues; and (iv) indirect social effects (illegal activities and fraud, alcohol control policy objectives or health aspects where applicable).

Market effects concern distortions of the quantity exchanged and of the equilibrium price of the various products. Taxation, by definition, distorts any market from the equilibrium that it would reach based on the free adjustment of demand and supply. For this reason, the present impact analysis does not assess market distortions per se, but those that might go beyond the intended objectives of the legislator. Conversely, what the analysis does take into account are aspects such as (1) tax-induced substitution across products, (2) cross-border distortions and illicit trade, (3) Single Market functioning in terms of possible distortions induced by diverging legal treatments, uneven application of the Directive or other administrative obstacles, (4) SME competitiveness since certain impact may have a differential effects on small producers vs. large manufacturers.

Regulatory costs and savings concern the broadly understood compliance, enforcement and administrative costs and cost savings. Compliance costs have been considered with respect to the changes to business practices linked to the administrative requirements. Enforcement costs and benefits can either relate directly to the costs borne by public authorities to apply the revised Directive provisions, or judicial costs and cost savings borne by public authorities and economic operators related to the need to interpret unclear legal provisions and, in case of judicial disputes, uphold them in court, as well as benefits (cost savings) in case interpretations and judicial disputes are no longer needed after a clarification or legal revision.

Tax revenues comprise direct charges including taxes and fees paid by economic operators or consumers. By nature, tax revenues bear elements of trade-off: what is a benefit for tax authorities is a cost for consumers and/or manufacturers. In the assessment and comparison of policy scenarios these impacts where primarily examined from the perspective of tax authorities. Impacts on tax revenues can be triggered, apart from the tax rates which are not part of this analysis, by scope of the tax system (exemptions / inclusions) and of individual tax category,

26 Regulation (EU) No 1169/2011 of the European Parliament and of the Council of 25 October 2011 on the provision of food information to consumers, amending Regulations (EC) No 1924/2006 and (EC) No 1925/2006 of the European Parliament and of the Council, and repealing Commission Directive 87/250/EEC, Council Directive 90/496/EEC, Commission Directive 1999/10/EC, Directive 2000/13/EC of the European Parliament and of the Council, Commission Directives 2002/67/EC and 2008/5/EC and Commission Regulation (EC) No 608/2004 Text with EEA relevance.

with the possible re-classification of certain products in different categories. It is also worth mentioning that these variations may also trigger other impacts, considered under market or social effects, such as tax-induced substitution between products, per capita consumption effects, demand for illicit products and fraud.

Indirect social effects include impacts that poorly lend themselves to quantification in monetary terms, but are nonetheless important since they concern the underlying values and principles of policy action that are linked to social well-being in broad sense. Two areas of social impact that have been considered related to the policy options at stake - although indirectly - namely: (i) public health (through alcohol control policy and measures); and (ii) tax fraud.

These broad impact categories constitute the general framework for impact analysis. Keeping in mind the complexity of the problem definition, the relative independence of the problem areas with distinct drivers, consequences and corresponding objectives, it should be recalled that not all of the impacts will materialise for all the problem areas and proposed options. For example, SME competitiveness is relevant but to the problem of reduced rates for small producers, the creation of a new fiscal category for certain product may generate administrative costs for economic operators who have to update their licenses and IT systems while health aspects are of relevance solely for the problem of denatured alcohol. However, where impacts might have been relevant but their nature cannot be defined (e.g. impacts on the consumption rates of alcohol after re-classification or changes in the scope of reduced rates), it is clearly stated so.

6.1. Dysfunctions in the application of exemptions for denatured alcohol

Mutual recognition of CDA

Option 1 would reduce the remaining legal uncertainties surrounding mutual recognition of CDA. There would be a reduction in any remaining trade barriers and market distortions as any restrictive interpretation of mutual recognition by some MS would be eliminated. This option would have no impact on most businesses, as this would only codify the approach taken by most MS. Some positive impacts for businesses involved in cross-border trade of CDA may also result. The impact of this option are summarised further in Annex 17.

PDA formulations

Market effects

All options would reduce, albeit to a different extent, barriers to intra-EU trade due to the greater transparency and legal certainty. Option 2 would result in fairer competition between PDA producers and users in different MS. The impact of Option 3 is uncertain, while information sharing could lead to fewer disputes / barriers to trade, this is dependent on MS adopting more consistent rules / practices. Option 4 would ensure equal treatment of PDA for indirect uses across the EU.

Operating costs for business / conduct of business

Any increase in harmonisation would be beneficial for businesses that operate across the EU. Option 2 may increase the access to wider range of PDA formulations and enable cross-border businesses to use the same PDA formulation in all MS, which would result in cost savings. There would be less legal uncertainty, which would reduce the risk and costs of supplying PDA intra EU. However this option may negatively impact businesses whose current formulation is not on the harmonised list.

The impact of Option 3 is uncertain as confidence / capacity building measures may not necessarily translate into savings for PDA producers or users. However if consistent rules or practices were adopted this would lead to a reduction in costs for businesses.

Option 4 would result in cost savings for businesses using PDA in MS that do not exempt indirect uses of PDA. The enhanced legal certainty of this option would reduce the risk of potentially costly disputes in the future.

Enforcement costs

The development of a harmonised list of PDA (Option 2) would result in a significant investment of resources by MS and the Commission. While this would build on the work undertaken by the Fiscalis Project Group, this would still be a major commitment for all stakeholders. However a harmonised list would reduce the burden of customs laboratories in certain MS, where fraud with illicit surrogate alcohol is a significant problem.

Option 3 would have some costs for MS. This would be dependent on the frequency and intensity of information sharing events. In time this may result in reduce enforcement costs for MS, if consistent rules were applied across the EU.

The implementation of Option 4 would have no additional burden on MS, who currently exempt the indirect uses of PDA. However for the other MS, this would involve updating legislation, adopting standard procedures and familiarise staff with the new rules and guidelines, which could constitute a small one off cost for the national authorities.

Fiscal fraud / Public health

Option 2 would not eliminate 'weak' formulations of PDA, which is the main source of fiscal fraud and negative health impacts. Some impact may be possible if a strict list of PDA formulations was developed and MS adopted their approaches to risk assessment and / or require stronger evidence before authorising an additional formulation. Option 3 would have limited, if any, impact on the risk of PDA fraud. Option 4 is primarily a fraud prevention measure and the requirement to move products containing alcohol over a suggested limit in accordance with Chapter IV of Directive 2008/118/EC would give authorities an effective weapon in the fight against fraud. The impacts of the various options are summarised in Annex 17.

6.2. Dysfunctions in the classification of certain alcoholic beverages

Tax revenues

The reclassification of certain products would have direct repercussions on the tax revenue for MS. The magnitude depends on the actual rates applied and the equilibrium of two opposite effects:

 A tax yield per product unit increases when the reclassification is to a higher taxed category  A higher tax results in a higher price, which has a negative impact on demand.

Taxing borderline products under Art. 20 (Option 1) would result in a direct revenue loss of approximately EUR 126 million per annum. This is due to a reduction in demand triggered by higher prices reflecting the higher excise duties. The clarification of the term 'entirely fermented origin' would have a modest impact on tax revenues as few if any existing products would be reclassified. If a strict threshold was adopted, this may result in the taxation of AFC as ethyl alcohol.

Selecting Option 2 (differentiation in the OFB tax category) may lead to yet more direct tax revenue losses, approximately EUR 250 million per annum. However, this estimate reduces to EUR 35 million if borderline cider is kept out of the reclassification process.

On the other hand, drawing on the experience from the introduction of relatively heavy alcopop / premix taxes in FR and DE, the medium-term to long-term net revenue losses may be much smaller. The introduction of a new tax in FR and DE did indeed, as expected, lead to the market collapsing quickly, and in a short time period the tax yield dropped to very modest contributions. Businesses largely withdrew products from the market that had become too expensive for the consumers and invested in other new products. This was the case with spirits- based alcopops which were replaced by malt- and wine-based pre-mix drinks after the introduction of the alcopop tax. Assuming similar market behaviour would follow from the reclassification (and thus new taxes), we can expect a similar process: a tax-shock would eventually result in the substitution of the target products with other products that remained in the favourable tax categories. The expected change in the excise duty revenues would depend primarily on which other products would be consumed and their level of taxation.

The non-regulatory Options 3.a and 3.b would not differ in terms of the nature of the expected impacts on the tax revenues. Those being operational measures, their deployment – be it independently or in conjunction with Options 1 and 2 - may reduce the risk of new misclassifications and bridge tax losses through smoother transition to new tax categories.

Introducing an EU wide regulation of cider and other specific OFB (Option 4.a) would have limited effects on tax revenues unless accompanied by the corresponding fiscal measures. The tax effects would depend on the final definition of cider, perry and fruit wines and could be similar to that of Option 2. Introducing a new EPC for OFB (Option 4.b) would have no impact on tax revenues, however it would provide enhanced data for national authorities to understand the OFB market better for tax policy decisions.

Competition and market effects

It is apparent that the reclassification of certain products into a different tax category with a different excise duty rate would have an impact on the market size and trends. Various steps were undertaken to assess the impact, which are detailed in Annex 15.

Option 1 would affect primarily borderline IP with an estimated reduction in sales volumes of approximately 36%. The collapse of this market is primarily due to the introduction of a higher excise duty on products that in various MS enjoy a zero or very low excise duty. It is further impacted due to the fact that the demand for these products is very elastic, so consumers would likely switch to other cheaper products.

The sub policy option of clarifying the term 'entirely fermented origin' would primarily have an impact on the certainty and consistency of rules across MS, but only limited market effects (0.3%) since the addition of AFC is, for the main, already accepted.

Option 2 would particularly impact very low strength mixed drinks and borderline cider, if included in the reclassification. The analysis estimates a decrease of between 46% (average scenario for very low-strength mixed drink) and 64% (average scenario for ‘borderline’ cider) in sales volumes. A moderate impact is expected for mixed drinks between 5.5% and 10% as these products are currently taxed as IP in some MS.

Both options may result in unintended effects on non-target products. Some AWP classified as CN 2206 may fall within the reclassification. Unintended effects are more profound under Option 2.

Overall both options have significant market impacts for the target products, since their demand is sensitive to price. The estimated decline in sales is substantial; however this is small when compared to the overall alcoholic beverage markets (less than 0.4% in the worst scenario). Similar impacts would be seen for the non-regulatory options, which aim to clarify the conditions under which certain fermented beverages should be treated like spirits.

Like above, the non-regulatory Options 3.a and 3.b would not differ in terms of the nature of the expected impacts on the tax revenues. Those being operational measures, their deployment – be it independently or in conjunction with Options 1 and 2 - may reduce the risk of new misclassifications and bridge tax losses through smoother transition to new tax categories.

Introducing a sectoral definition of cider and other specific OFB will have limited impact on the market, if its introduction is not accompanied by the corresponding amendment of the Directive. If the Directive is amended, the market impacts are similar to that of Option 2.

Administrative burden and enforcement cost

The policy options can have an ambivalent impact on administrative costs and burdens for businesses and competent authorities. They intended to reduce the current burden caused by classification issues and uncertainties. However the introduction of new measures may result in additional costs for adapting existing systems and implementing new rules. Since there was no sufficient and reliable data to calculate the burdens in monetary terms, any quantification attempts were only possible on the basis of hypothetical scenarios.

Option 1 would not impose new costs for all stakeholders beside the 'one off' need to familiarise with the new rules and guidelines, adopt standard procedures and train staff accordingly. As for businesses, the staff efforts required to familiarise and implement the new rules may vary by company size. The affected population encompasses in principle all those who produce ‘borderline’ CN 2206 products, these can be found primarily among OFB producers, but also among certain breweries and wine/liqueurs producers. The Study estimates that the familiarisation costs would amount to approx. € 4,500 per company. These costs are likely to be supplemented by the costs potentially incurred to review the production processes, economic portfolios or market strategies. The impact on competent authorities could not be quantified in the Study. However, in terms of unit costs it would be expected to be higher although – given that the affected population is limited - in aggregated terms it may be modest. The Study estimates that the aggregated benefits would possibly offset costs within a 5 – 6 year period.

Indirectly, Option 1 may reduce the number of complex dossiers by ca. 50%. That would lead to an estimated costs savings may amount to some EUR 1 - 1.5 million per year, for competent authorities. Non-quantifiable benefits for businesses in the same proportion can be assumed.

Option 2 is less oriented toward ‘difficult-to-classify’ products, and the burden due to the difficult distinction between CN 2206 and CN 2208 would persist. Moreover, distinction may create new ‘borderline’ products which could, in the worst case scenario, neutralise benefits of any new clearer definitions. It can therefore be reasonably assumed that the overall present burden would not change significantly (EUR 2 – 2.5 million). Option 2 would require various administrative actions, including familiarisation with the new rules and guidelines (similar to the costs associated with Option 1), amendment of legislation, updating the IT systems, training of staff, updating some national procedures for licensing and authorisations, etc. All the required action would be ‘one-off’, no relevant recurrent cost is envisaged. The affected population includes primarily OFB producers and the IT adjustment associated with changing of the EPC systems are estimated to amount to approximately EUR 800 per economic operator (weighted by enterprise size). The costs of this option would be offset by the benefits in 10 years or more,

which is longer than option 1. Also in this case the impact on competent authorities cannot be quantified but the overall costs are expected to be modest.

Furthermore, there are some important considerations related to the choice of demarcation Line A or B for Option 2. The demarcation Line A would require the adoption at EU level of a harmonised definition of cider, perry and the other OFB that correspond to the original scope of this category and match as much as possible with existing national criteria for these products. This is far from being straightforward: national definition vary significantly, the industry calls for a permissive approach e.g. establishing no minimum amount of fresh juice, no limits to added sugar and water etc. – which is probably tantamount to shifting Line A to overlap with Line B; whereas certain consumers organisations consider most of the mass-market products not to be ‘real’ cider. It is apparent that the Directive is not the appropriate vehicle for product definition, which should instead be developed as sectoral legislation.

Analogically, the demarcation Line B would require the definition of what a mixed drink is and the relevant criteria to allow for such a categorisation. Also in this case various approaches exist and an agreement should be reached among MS at the expert group level. The French definition seems more all-catching than other mixed drinks definitions in that it applies either to mixture of different beverages or to beverages with a certain amount of added sugar/sweeteners. In this respect, it may encompass also various ‘mass-market’ ciders - that means Line B shifts leftward to nearly coincide with Line A.

The administrative burden of Option 3.a and 3.b would require efforts and resources in all phases of their development and implementation cycle. There is no precise estimate of the overall cost, but Options 3.a and 3.b would be in line with the expected costs and benefits of Option 1. As noted above Option 3.a is currently underway. The development of a sectoral regulation (option 4.a) for cider, perry and other specific OFB without the corresponding amendment in the Directive would have no benefit from a tax perspective. As a complimentary measure, it would result in similar cost / benefit to that of Option 2.

The costs of a new EPC (Option 4.b) include the update of the existing excise systems used by both businesses and MS. The change envisaged is minimal, however all IT systems, templates, manuals etc. should be updated to include the new EPC. The administrative burden for authorities is possibly greater and involves the amendment of regulation and standard operating procedures, informing and training businesses at all levels, and obviously the direct costs of updating the IT systems. The unit cost per MS would vary in accordance with the specificities of the administrative system in place and the size of the country, but interviewees were not able to provide a quantitative estimate.

While there are costs associated with the introduction of a separate EPC for OFB, its introduction would bring significant added value in terms of monitoring and control of the market and excise duty trends. Currently tax authorities are seldom able to differentiate, and therefore to appreciate the market trends of OFB, which is the category that mostly contains new and ‘borderline’ products, so they have access to limited data evidence to support their tax policy decisions.

In the event of further changes of the excise duty structure, such as new tax categories to differentiate among OFB, the revision of EPC would become necessary for a proper management and monitoring of products movements, so this option would become justified also in costs/benefit terms. The impacts of the various options are summarised in Annex 17.

6.3. Dysfunctional application of reduced rates

6.3.1. Option 1 – Improve the functioning of reduced rates for small breweries

Competition and market effects

As described in the baseline analysis, the small brewers market is growing at a very fast pace and most likely the frequency of cross-border trade will increase. It is logical to presume that, left unfixed, the dysfunctional application of the reduced rates for small brewers will increase and may lead to unfair competition in the single market.

The clarification of the conditions at which a small brewer shall be considered independent will benefit the public authorities called to implement these provisions, as well as to small brewers. Indeed, should this clarification be introduced, it would be easier for public authorities and businesses to determine whether certain business models or decisions are compatible with the reduced rate schemes.

For small brewers, this would reduce the risks connected to the entering into certain trade relationships, as well as the litigation costs associated with cases where the interpretation of the customs authorities will be challenged by the operator. Also, the discrepancies between MS or between regions of the same MS – which have been sporadically reported – will be tackled, reducing the risk of an uneven treatment of similar situations.

An improvement in the legal clarity of the provision for cooperating breweries, and a smoothing of the procedures for intra-EU trade are a positive factor for the competitiveness of SMEs. In particular, this would benefit larger players across the SME population, which are more likely to enter into cross-border trade or into more complex contractual relations, favouring their business growth. At the same time, increased ease of doing business for intra-EU traders could have a positive market effect for cross-border businesses, and eventually result in an increase of intra- EU trade flows. However, the scale of the problem at stake is modest, meaning that the procedures to apply the reduced rates do not represent a high barrier to the functioning of the single market. Hence, benefits are likely to be modest.

Administrative burdens and enforcement costs

Any clarification to how reduced rates should be applied to businesses established in a different country than that in which the beer is released for consumption would affect the administrative burdens borne by businesses and the enforcement costs borne by public authorities.

Under Option 1.b, a uniform certificate issued by customs authorities upon request to any EU brewer could serve as a means of proving the status of small brewer. Such a certificate could be designed at EU level, included in a binding norm, and would be accepted by all customs authorities in the MS of destination. Such a certificate should provide information on the annual output and the independent status of the brewer.

Under this approach, companies which are already small brewers under national rules would incur limited administrative or enforcement costs (i.e. the costs of requesting the certificate). Total burdens for the 675 operators in the sample MS would amount to approximately EUR 13 000 or 2% of the burdens estimated for the overall scheme. The situation would be different for businesses who are not small businesses in their country. Therein, to claim the reduced rate, the brewer would need to prove his/her status as an independent economic operator, by submitting the customs authority the required documents (company registration, information on shareholding, company charter etc.). Administrative burdens for the 180 operators not under the

scheme are estimated at EUR 32 000 or 5% of the burdens estimated for the overall reduced rates schemes. Overall, the additional administrative burdens seem limited for this policy option.

Enforcement costs for public authorities are considered to be modest, when dealing with businesses already benefitting from reduced rates in their country of establishment. There would be some further administrative burdens for businesses and enforcement costs for public authorities as a legislative revision would be needed to introduce a uniform certificate, so that the format and content of the document could be fully harmonised at EU level. The impacts of the various options are summarised in Annex 17.

6.3.2. Option 2 – Extending the reduced rates to small cider makers

Competition and market effects In terms of market competition, small cider makers would gain relatively to large ones, either because they are able to reduce their price or increase their profit margins (or a combination of both). The reduction compensate for higher costs of production due to diseconomies of scale, which mirror those suffered by small brewers. The sheer difference in size between industrial producers and small cider makers, and the very small market share retained by the latter imply that reduced rates would hardly represent a significant competitive threat for large players.

The competitiveness of SMEs in the cider industry would be enhanced by the provision. Impacts could be estimated to be analogous to those enjoyed by small breweries, given the similarities in terms of market structures.

Administrative burdens and enforcement costs As far as administrative burdens are concerned, it is assumed that the annual burdens per small cider maker would be similar to those incurred by small brewers, estimated at EUR 178. The EU population potentially covered by the provision is estimated at about 1 145 small cider makers.27 Total burdens are thus estimated at about EUR 200 000. Considering the market share of small cider makers in MS applying a positive tax rate, and thus potentially affected by the provision, costs per unit of production would amount to 0.32 EUR/hl.

Finally, in terms of enforcement costs, public authorities would have to deal with a new scheme, and thus with the associated demands to obtain the reduced rates. This would engender additional costs, but the number of players at stake is so limited that those costs would not be large. Extra-EU imports of cider represent 0.1% or less of EU consumption and only a share of that might be produced by small cider makers; hence no significant hurdle is expected in the management of possible applications from non-EU small suppliers.

Tax revenues Forgone tax revenues are unevenly distributed due to the dimensions of the EU cider market. MS where the cider market is large and the excise duties are high, such as IE and the UK, the total forgone tax revenues based on a 50% reduced rate for small cider makers are estimated to be EUR 1.3 million and 9.7 million, respectively. Impacts are estimated to be less than EUR 0.5 million in MS such as PL and FR with small cider markets and low excise duties.

27 Based on the estimated number of small producers in the 5 MS and their share of consumption over total EU consumption

Health impacts for consumers The effects on per capita alcohol consumption, and consequently health impacts, are expected to be negligible. The portion affected by the extension, estimated at 4.6% of the cider market, is too small to affect the overall price and consumption of cider. In addition, cider represents a relatively smaller market compared to other alcoholic beverages in most of the MS. Only countries with a very large cider market, the UK and IE, could see noticeable negative health effects, if the reduction was introduced. The impacts are summarised in Annex 17.

6.3.3. Option 3 – revised threshold for low strength beer

This option is expected to generate impacts in terms of: (i) tax revenues, as larger shares of the market could benefit from reduced rates compared to the baseline; (ii) market effects, as lower taxation may lead to lower price for low-strength beer, hence an increase in demand; and (iii) ambivalent public health effects, as increased consumption of low-strength beer may (or may not) reduce the per capita intake of pure alcohol and, through higher availability, may increase the number of alcohol consumers, particularly among price sensitive consumers such as young people, heavy drinkers and people from lower socioeconomic groups.

Competition and market effects It is apparent that the market share for low-strength beer between 2.8% vol and 3.5% vol across all MS is modest; nonetheless, it is reasonable to assume that the adoption of the 3.5% threshold would develop a new ‘niche’ market immediately below this limit.

Tax revenues The total foregone tax revenues (including VAT paid on excise duty) are expected to amount to less than 1% of the total tax revenue from consumption of beer in the selected MS. Foregone tax revenues might be even lower, if one considers that the new market for low-strength beer could partially flourish on top rather than at the expenses of the market for standard beer.

Health impacts for consumers Considering the above analysis of market effects, and more specifically the possible limited increase in per capita consumption of low-strength beer (from 0.02L to 0.10L per year), any public health impact, either positive (where the additional consumption of low-alcohol beer is ‘at the expense’ of standard beer and other stronger alcoholic beverages) or negative (where low-strength beer substitutes soft drinks, or increase the overall consumption of alcoholic beverages and facilitates the drinking initiation of young people), can be considered negligible. The impacts are summarised in Annex 17.

6.4. Unclear provisions to measure of Plato degree for sweetened / flavoured beer

As discussed above, all options and sub-options (except the baseline) revolve around the selection of one of the three existing approaches to measuring Plato degree and, therefore, they would have the same type but not magnitude of impact. The analysis presented here is based on the sample of 6 countries selected for the case studies under the Study (AT, BE, DE, IT, PL and RO). These countries represent the large majority of the sweetened/flavoured beer market in the EU countries that have adopted the ‘Plato’ method.

Tax revenues and market effects As described in the baseline analysis, the EU market for sweetened/flavoured beer is expected to grow fast in the coming decade and most likely the frequency of cross-border trade will increase. Whereas precise estimates are not available, it is logical to presume that unfixed, the

problem may lead to unfair competition in the domestic and single markets if alcoholic strength is calculated based on different methods.

It should be noted that the present appreciation of impacts on tax revenues and market effects has a domestic market angle. Impacts are country-specific and depend not only on the approach applied to measure Plato degree of sweetened/flavoured beer, but also aspects such as the level of excise duty, VAT and market segments.

Selecting approach A or B1 would result in an overall reduction in tax revenues (excise duty and VAT on excise duty) from sweetened/flavoured beer of more than EUR 30 million (about - 25%), compared to the baseline situation. Consumption, on the other hand, might increase by approximately 100 000 hl in the 6 countries combined, i.e. less than 2% of the total consumption of sweetened/flavoured beer. Limited changes in consumption reflect limited changes in prices. Selecting approach B2 would result in minor changes as opposed to the baseline approach as this is the approach currently in force in most of MS considered. Expressing the changes as a percentage of the total beer market, the impacts become rather negligible: between +0.2% (moving to approach B2) to -1% (selecting approach A or B1) for tax revenues (including VAT on excise duty), and between almost nil (selecting approach B2) to +0.1% (selecting approach A or B1) for consumption volume.

Different MS will have different baseline scenarios, depending on which method they currently use and to which method they would need to switch (see Annex 14). As approach B2 generates the highest excise revenues, countries that will need to change away from this method are likely to experience some revenue decrease (as it is confirmed by the analysis of DE, AT and to a lesser extent PL and BE). Countries currently using another method than B2 would see a sharp increase in the excise revenues, for example in RO.

Analogous patterns are also observable for the price and consumption changes. Countries which would decrease their excisable tax base (discarding approach B2) should expect a drop in price for sweetened/flavoured followed by corresponding increased consumption. In countries like RO, this effect would be reversed.

While approaches A and B1 lead to similar value of the Plato degree of sweetened/flavoured beer and somehow reflect its actual alcohol strength, approach B2 leads to higher Plato degree, possibly greater than the Plato degree of a standard beer with an equivalent alcoholic strength. For instance, approach B2 results in almost double the Plato degree of a typical radler when compared to approach A or B1.

In principle, approach B2 is therefore more prone to generate possible distortion of competition between standard and sweetened/flavoured beer. However, as the impact analysis showed, the actual changes in price level that can be expected from switching between different approaches are rather modest, and of limited importance vis-à-vis other competitiveness factors. Overall there is a negligible risk of an excessive market distortion caused by the selection of any of approaches.

Public health Any significant impacts potentially stemming from the harmonised adoption of any of the three approaches considered appear to be unlikely or limited. In fact, based on Eurostat date for total population above 15 years, the annual average per capita consumption of sweetened/flavoured beer in the six surveyed MS would range from 2.67 litres per annum (selecting approach B2) to 2.73 litres per annum (selecting approach A or B1). The difference is clearly negligible when compared to average per capita consumption of ‘traditional’ beer, which in sample MS exceeds 78 litres per annum.

Enforcement and legal costs When it comes to enforcement, any change in current approaches would require some MS to adapt their monitoring and control procedures. As mentioned, approach B2 is the most used, so the overall number of countries that would have to modify their systems would be limited when approach B2 is defined as most appropriate. Moreover, approach B2 allows authorities to perform checks directly on the end-products, with no need for on-site inspections and/or measurement during the production process, and is therefore considered more cost-effective than the other approaches. For these reasons, the selection of approach B2 at EU-level would have little or neutral effect on the enforcement costs for MS authorities.

Conversely, the customs authorities interviewed explained that, as things now stand, it is not possible to compute the parameters required to apply approach A or B1 by analysing the bottled ‘end-product’, since the current analytical methods do not allow for it. Therefore, the enforcement of approaches A and B1 would require checks at the production facilities, and these may generate new one-off costs, such as the devising of operational rules and the installation of measurement equipment, as well as recurring costs in the form of on-site inspections. An additional issue concerns sweetened/flavoured beer produced in another MS or third country, since the authority of the MS where the product is released for consumption could not directly conduct inspections and would be reliant on the information provided by the businesses and/or, in certain circumstances, by the authority of the producing country.

Finally, the selection of a harmonised approach to measure the Plato degree of sweetened/flavoured beer would increase legal certainty and eventually reduce the risk of disputes between tax authorities and brewers. All impacts are summarised in Annex 17.

7. HOW DO THE OPTIONS COMPARE?

As regularly recalled, the issues at stake in the present initiative are relatively independent from one another. Therefore, the comparison of options has been performed for each thematic area separately, rather than in a cumulative way. For the sake of transparency and clarity, all objectives are considered in the analysis of effectiveness even though some of the options were never designed to meet them. However, care was taken to ensure that all of the options are at least neutral (no impact) towards any of the objective.

This is reflected in the comparison table at the end of this section, while the narrative of the analysis focuses only on the objectives and impacts relevant to the particular policy option.

7.1. Dysfunctions in the application of exemptions for denatured alcohol

7.1.1. Comparison of options

Mutual recognition of CDA

A regulatory amendment (Option 1) of the Directive will ensure that divergent interpretations involving MS that have notified CDA formulations other than the Eurodenaturant will be eliminated and legal certainty will be achieved.

This option is in line with the approach with most MS and as a result it will have little impact on tax revenues of MS and it will not increase costs for businesses. For the main this option is codifying the existing practice. This option will reduce any remaining trade barriers and distortions and consolidates MS desires for a harmonised solution for CDA into a legal text.

PDA formulations Effectiveness The extent to which Option 2 or Option 3 would effectively meet the policy objective of legal certainty is limited. Option 2 would increase the transparency and certainty surrounding PDA formulations, however there is no guarantee that legal certainty would be achieved. While the list would be agreed by all MS, MS would retain flexibility to authorise other formulations in cases where the fiscal risk is demonstrably low. This concept of low fiscal risk currently varies significantly between MS and the possibility to authorise other formulations limits the transparency.

Option 3 is effectively a complimentary measure and would be ineffective in creating legal certainty. Option 4 would enhance the clarity surrounding the legal meaning and uses of PDA. This would eliminate ambiguity and uncertainty that currently exists in relation to PDA.

Efficiency Option 2 would result in fairer competition between businesses in different MS, however the costs for MS and the Commission would be significant. These costs would be balanced by the savings / benefits of a harmonised lists, which would reduce the workloads of custom laboratories and the risks associated with cross-border trade that currently exist for businesses. The confidence / capacity building measures of Option 3 would be efficient in terms of increasing the trust between MS, however as an independent option, the overall efficiency is highly uncertain.

Option 4 would be efficient as the costs of clarifying the legal base for PDA would result in benefits for businesses in terms of legal certainty. This would ensure equal treatment of goods containing PDA across the EU and reduce the risk of costs associated with disputes between businesses and national authorities.

Coherence As noted above Option 2 and 4 would result in improving the functioning of the single market. Option 3 may assist in increasing information flows between MS, however the overall coherence with other EU policy objectives is highly uncertain.

7.1.2. Stakeholders views

Mutual recognition of CDA

Most stakeholders interviewed as part of the Study, as well as a small majority of respondents to the OPC, were in favour of the harmonisation of CDA formulations. However there was strong opposition from a limited number of MS. The response to the OPC attracted a low response level and for the main a neutral response was adopted. In the case of continued uncertainty regarding the mutual recognition of CDA, 41% (38 respondents) agreed that the continued use of national formulations causes legal uncertainty, with only 7% disagreeing.

PDA formulations

The development of a harmonised list for PDA was strongly opposed (73%) by industry stakeholders with an interest in the production or end use of industrial alcohol. Overall a small majority (51%) of respondents to the OPC disagreed with this option.

The industry also expressed a strong disagreement with a strict interpretation of the legal base for PDA formulations. Instead respondents (85%) supported capacity and confidence building measures in order to improve the understanding of MS' approaches.

7.1.3. Comparison summary and preferred option/package of options

3 – Option 2 – PDA 4 –PDA No 1 - CDA capacity list terms change EFFECTIVENESS building ensuring fair treatment of businesses ++ + 0 ++ 0 across all alcohol sectors preventing and correcting any ++ + 0 ++ 0 distortions of competition providing clear rules on the scope, classification and calculation of duties ++ + 0 ++ 0 for businesses and MS providing clear and efficient conditions to determine denaturation ++ + 0 ++ 0 procedures reducing administrative burden and compliance costs for businesses and 0 0 - + 0 tax authorities provide legal certainty ++ 0 0 ++ 0 strengthening the fight against fraud + + + ++ 0 and tax evasion improving human health protection 0 0 0 + 0 EFFICIENCY administrative burden 0 0 - + 0 tax revenues 0 0 0 0 0 COHERENCE ++ + 0 + 0 OVERALL ++ 0 0 ++ STAKEHOLDERS OPINION + - ++ -

In terms of CDA, Option 1 of amending the Directive to clarify the mutual recognition of CDA (hybrid recognition) is the only and the preferred option to ensure legal certainty within this area. For the record, alternative modalities of improving the mutual recognition were analysed and discarded early on in the process.

The preferred option in terms of PDA is Option 4 to clarify the unclear wording of the Directive to increase the legal certainty for its indirect uses and finished product containing PDA. The capacity / confidence building measures under Option 3 is also an option that is worthwhile, however this option will be complimentary as its success as a standalone approach would be mininal.

The package of options under the cluster of measure relating to the treatment of denatured alcohol is therefore composed of the bundle of Option 1 + Option 4 accompanied by Option 3, on a complementary basis.

7.2. Dysfunctions in the classification of certain alcoholic beverages

7.2.1. Comparison of options

Effectiveness In terms of legal costs, the overarching rationale for all options is to reduce legal uncertainties and disparities of interpretations of certain products. The effectiveness of the various options appears uneven with not one option achieving this without negative impacts.

Option 1 would reduce the disparities of tax treatment of similar products as the classification for excise purposes would not be so strictly determined by CN codes. Instead the classification would also be linked to the CJEU rulings in this area. This option alone, as it was flagged out by stakeholders (especially in tax administrations), would not bring the desired effects as the current uncertainties in the underlying CN classification would persist. Due to the subjective nature of the CN explanatory note and the CJEU principles leaving ample room for interpretation, there would be a need for robust guidelines on the conditions, criteria and methods to treat the borderlines products (offered by options under cluster 3). For the internal market to correct the discrepancy, it would be most effective across the EU if customs classifications also took account of the CJEU criteria, which is currently underway (Option 1 + Option 3.a/3.b).

The final element of Option 1 which involves clarifying the term 'entirely fermented origin' in relation to AFC would remove the degree of uncertainty that the current ‘patchwork’ of national solutions inevitably cause, which may create unnecessary hurdles and delays in operations and eventually constrain the full deployment of the market potential.

Options 2.a and 2.b would result in legal certainty at EU level and a consistent treatment of borderline products across MS since it would make the current national level non-harmonised distinctions unnecessary. As with Option 1, Options 2.a and 2.b would be most effective if accompanied by robust definition for the new category, with available guidance on conditions and criteria allowing classification under that or another tax code (Option 2.a/2.b + Option 3.a/3.b). At the same time, implementing new tax category would necessarily impose an administrative cost and burden on businesses and tax authorities as they would need to review their existing national excise duty systems from a legal and technical perspective, which is further analysed under efficiency.

Under approach 2.b (demarcation line at borderline products of mixed drinks), in order to avoid competitive distortions, the structure (and level) of taxation would most likely be in line with that applicable to ethyl alcohol. It would however increase the complexity of the excise law and create incongruity for the EMCS, which does not distinguish between OFBs (traditionally closer to wine) and wine. Resolving the EMCS would not only be costly but also undesirable as the original intention of the OFB category was to protect other traditional products of fermented origin – for example cider and perry - from higher taxation. For these reasons, in order to avoid instituting differences in the tax category which may unintentionally exclude some eligible products, erode this tax category in legal terms and increase the overall complexity of the system, the preferred approach is the approach 2.a distinguishing for tax purposes traditional cider and perry from all other OFB.

This will enable MS to introduce such differentiation into the OFB tax category and enable them to apply different excise rates to these products if so desired. Furthermore this differentiation will ensure the application of reduced rates (see section 5.4.3) is restricted to (small) cider and perry makers. Put differently, it is considered more effective to increase legal certainty to sub-

define the OFB products that lend themselves to more distinct definition without deeper fragmentation of this tax category.

Options 3.a and 3.b – although viable on their own – would alone yield equally uncertain benefits albeit for different reasons. As for Option 3.a, its strength lies in mitigating the negative results of the current CN code uncertainties which would thus not be replicated to the excise duty level, and the risk of more severe legal disputes may be avoided. Revision of the CNEN towards closer correspondence in the interpretation of the CJEU principles in both excise and customs classification, would therefore eliminate the very source of disparities, being thus very effective. Option 3.b, consisting of non-binding guidelines, would necessarily leave a certain room of interpretation to MS authorities. Therefore, this sub-option would be comparatively less effective in ensuring a harmonised treatment of the same products across different MS. The BTIs would no longer constrain the tax categorisation and their use would likely reduce, but the absence of this practical instrument may eventually trigger the perception of a higher degree of uncertainty and unpredictability by businesses. Moreover, the risk of non- robust definitions or non-compliance (given the non-mandatory nature of the guidelines) may constrain the effectiveness. If the tax categorisation remains determined by the unchanged CN codes, non-harmonised national measures for special products may persist or even accelerate.

Depending on the scope of Option 3.a and 3.b, it could be argued that at best, they could together pre-empt the need to amend the Directive and would also result in sufficient legal certainty at EU level. However, as these options are outside the Directive, this would require the involvement and consensus of several different services of the national and European administrations, which will naturally impact negatively the efficiency of its implementation.

Option 4.a involves the adoption of a sectoral definition, which would assist when categorising OFB within the Directive, however as a standalone option, it would not address the current problem of different classifications of alcoholic products. An amendment to the Directive would still be necessary.

Option 4.b would enable tax authorities to enhance the data they currently receive through the existing excise systems. This would improve their tax policy decisions.

Efficiency Option 1 would transpose the CJEU rulings into the Directive and would impose 'one off' minimal costs and burdens on businesses and national administrations. Due to the subjective nature of the CJEU criteria, the overall efficiency of this option is questionable, as variances between MS will persist and disputes may continue.

As briefly mentioned under the effectiveness criterion, implementation of Option 2.a and 2.b would trigger adjustment costs and burdens to businesses and national administrations alike, stemming from separating the category into cider/perry from other OFB. This is because there seems to be (i) relevant disparities in the legal definitions that already exist in the different MS, which should be aligned; and (ii) diverging views between producers of ‘mass-market’ products and their trade associations, and small ‘traditional’ producers and certain consumers’ organisations.

Furthermore, with some exceptions (e.g. IE, UK) these products are typically regulated in national food and agriculture legislation, so the Directive does not seem to be the most appropriate vehicle for establishing a common product definition. At the same time, there might be some rationale to pursue an EU-level definition of cider etc. outside of its fiscal treatment. Cider has historically never been clearly defined in its own right - it follows (along with OFB) the rules on rates for wine. As the cider industry has developed, there has become a need for a

more efficient structures regime to define cider (& perry, fruit wines and mead) separately within the category of OFB especially mixed products.

In the case of mixed products, the main challenge would consist in adopting a definition that does not simply create tax incentives to develop substitute products, as it happened for instance with the ‘alcopop’ tax in Germany. On the scope of this category, MS may have different views related to the specificities of the national industry and market and might want to include or not malt-based mixed beverages and so called ‘wine-coolers’.

Option 2.a would be more efficient and easier to implement than option 2.b. Adopting a definition for alcopops runs the risk of creating a new tax incentive to develop substitute products. This would result in further amendments to take account of future developments. Cider and perry are traditional products and the basis of their production remains the fermentation of apples and pears.

Implementation of Options 3.a and 3.b, as they fall outside of the remit of the excise duty system, would require a larger consensus at the international level, in order to avoid any hurdles and uncertainty affecting the international trade. Option 3.a is currently underway and expected to be completed by June 2018 and will complement the final option chosen.

Both options may result in reductions in demand for borderline products, which would negatively impact tax revenues of MS and the changes analysed would likely not lead to beneficial effects. Furthermore, the only benefits would come from products that would be unintendedly affected (e.g. AWP). There is some reasoned expectation that consumers' preferences would largely shift to other alcoholic beverages, so the net tax loss would be mitigated. Overall however, a minor tax loss can be expected, since the main alternatives to borderline products are more lightly taxed.

All options would have the recurring benefit of a reduction of administrative burden and would involve one off costs. As a result the balance of costs and benefits would shift over time. The costs of Option 1 would be offset within 5 – 6 years, whereas Option 2 would take longer (10 years or more). Options 3.a and 3.b require more effort and resources of more stakeholders and therefore it is difficult to estimate the balance of costs and benefits of these options beyond a reasoned assumption that their implementation could be significantly hampered.

Coherence All policy options have competition and market effects. Option 1 would negatively impact the demand for borderline products and may impact non target products unintentionally. Furthermore in the absence of robust criteria this could have a severe impact on trade as the BTI tool would no longer ensure the same tax treatment of a product across the EU (including imported products). Option 2 would increase harmonisation across the EU market but like option 1 may unintentionally impact non target products. Option 3 would be effective as the uncertainties relate to customs classification and the differentiation of EPC and these options would improve the functioning of the internal market.

All options may be seen through the eyes of stakeholders as incoherent in terms of correspondence with the present national legislation or practice. Defining a product definition or providing guidance on classification will inevitably impact some countries more than others, depending on which approaches and definitions are chosen. However, given that the lack of coherence in applying tax and customs treatment to the products broadly classified under OFB is the very problem at stake behind the present initiative, the coherence aspects should be seen from the perspective of the single market. In that case all options are coherent with the objective of ensuring smooth functioning of the internal market and ensuring coherence of product

treatment in each geographical market. The difference will lie in the effectiveness and efficiency with which this sought for coherence will be achieved, which distinction is duly analysed under the two other respective comparison criteria.

All options are also broadly coherent with the Council conclusions calling for the necessity to prevent ambiguities leading to distortions of competition between businesses and to apply harmonised conditions and rules for taxing alcohol and alcoholic beverages. The Council specifically recognised the need to clarify and to harmonise further the classification rules for products manufactured as mixtures of different categories of alcoholic beverages or as mixtures with non-alcoholic beverages or OFB in order to unify the treatment for excise purposes of the same products across the MS, and so ensure legal certainty and clarity for businesses.

Option 1 would be coherent with the CJEU rulings and is likely to be less disruptive to MS as this is the current criteria used by MS to classify these products. However the current work on the CN codes (Option 3.a) would also need to be incorporated into Option 1 to ensure consistency of approach. Defining cider, perry and fruit wines (Option 2.a) would be in line with other alcoholic beverages such as wine and spirits, which have sectoral definitions. Aligning a sectoral definition with a new category for traditional OFB would ensure coherent across EU legislation (option 4.a).

7.2.2. Stakeholders views

The level of agreement between the OPC respondents is mixed and can be easily related to the perspective of specific segments of the industry and / or interest of other nature. Respondents often conceded that there can be added value in a general clarification of the current situation, however they believe that the perceived risks of a legislative change tend to outweigh the perceived benefits across all respondent groups with the exception of private individuals. A clear majority of industry respondents believe that a revision of the OFB tax category would generate negative effects on all fronts, including adverse effects on international trade, classification uncertainties and disputes and market distortions.

Almost half of respondents (48%) agree that beverages like cider and perry should be defined separately and not under the generic OFB label (24% disagreed with the option). This increases to 53% of stakeholders with an interest in the cider sector with the remainder neutral and to 68% of private individuals.

In terms of the approach to the classification of certain alcoholic beverages, 68% of respondents agreed, if not strongly agreed with incorporating relevant parts of CJEU judgements into the Directive. The option of creating a new category for cider, perry and fruit wine was positively received by the beer and cider industries (56% and 64% respectively) but only 35% and 38% of wine and spirits producers agreed with this option. However private individuals strongly supported the new category with 69% in favour of this option.

73% of respondents to the OPC would like the meaning of the concept of 'entirely of fermented origin' clarified so as to define the status of products containing AFC, with only the spirits industry expressing a more cautious opinion. A mixed response was received in relation to non- regulatory options.

41% of respondents to the OPC supported the amendment to the EPC to separate OFB from wine. A further 29% expressed a neutral position, while 30% of respondents disagreed.

7.2.3. Comparison summary and preferred option/package of options

Option 1 – 2a– new 2b– new 3a – CN 3b – non 4a – 4b - EPC No

CJEU sub sub EN binding sectoral change* rulings category category guideline definitio EFFECTIVENESS for for other s n cider/ mixed perry drinks ensuring fair treatment of businesses across all alcohol - ++ + - - - 0 0 sectors preventing and correcting any - + + 0 - - 0 0 distortions of competition providing clear rules on the scope, classification and - + - - - - 0 0 calculation of duties for businesses and MS providing clear and efficient conditions to determine n/a denaturation procedures reducing administrative burden and compliance costs for 0 - - - 0 + 0 - 0 businesses and tax authorities provide legal certainty - ++ + 0 0 0 0 0 strengthening the fight against n/a fraud and tax evasion improving human health n/a protection EFFICIENCY administrative burden 0 - - - 0 + 0 - 0 tax revenues 0 - - - - - 0 0 COHERENCE + + + ++ ++ + + 0 OVERALL 0 ++ + 0 0 0 0 STAKEHOLDERS ++ + 0 0 0 + + OPINION

The preferred option whose deployment would be crucial to achieve the objectives is Option 2.a splitting the OFB category into two subcategories of which one would maintain the current treatment, while the other would ideally comprise of all traditional OFB products (i.e. cider and perry etc.) which would be defined and treated separately. While this option has downsides, including increased burden on businesses and tax authorities, this is the preferred option as this would reduce the disparities of treatment of similar products and would ensure the effective operation of the reduced rates for small cider makers (if introduced). Work is currently underway in improving the CN explanatory notes (Option 3.a), which as a complimentary option would assist in reducing classification disparities. Therefore, this approach alongside the non binding guidelines under Option 3.b could also form part of the preferred option package as they can improve the overall effectiveness of the functioning of the OFB category through providing operational definitions, criteria and methods, irrespective of what changes to this category will have been made. In other words, Options under cluster 3 would work just as well with Option 1 as with Option 2 or independently (albeit less effectively) and there is no reason to not include them in the preferred package. Options under cluster 4 could also form part of the preferred package going forward. Also these are complimentary options which would improve the functioning of Option 2.

The package of options under the cluster of measure relating to the classification issus is therefore composed of the main Option 2.a accompanied by Option 3.a /Option 3.b/Option 4.a/Option 4.b on a complementary basis.

7.3. Dysfunctional application of reduced rates

As the problems related to the application of reduced rates are multifaceted and independent from one another, the options are compared in sub-clusters related to the specific problems, following as well the logic of the presentation of impacts earlier on.

7.3.1. Comparison of options

Options under cluster 1: Improve the functioning of reduced rates for small breweries Effectiveness In terms of legal costs, the overarching rationale for both options and sub-options is to reduce the legal uncertainties and disparities of interpretations of 'legally and economically independent' and to improve the cross-border functioning of the scheme. The choice of the means to introduce this policy option – hard versus soft law – would have impacts over the level of legal certainty achieved.

Option 1.a.1 would involve a regulatory amendment of the Directive, which will ensure the policy objective of legal certainty is achieved. However due to the fast changing industry, this may result in a definition becoming obsolete with new developments. Option 1.a.2 would allow for a degree of subjectivity, which could quickly address any new market developments. This would enable MS to resolve any new issues without resorting to binding legislation, based on the consensus of national authorities. However MS would retain the power to apply it or not. The creation of an uniform certificate (option 1.b) for recognising small brewers would be a regulatory amendment and therefore ensuring legal certainty for businesses.

Efficiency Currently the reduced rates for small brewers works well for the main and does not generate unnecessary administrative burdens or enforcement costs. The clarification of the term 'legally and economically independent' would not result in any increased costs for the various stakeholders but would improve the overall efficiency of the relief. The verification of small brewers would have some administrative burdens or enforcement costs for ecomomic operators or public authoritites. The development of a certificate would result in a small increase in administrative burdens for ecomomic operators (estimated at 7.5% of total burdens from the scheme). Public authorities would incur modest additional costs, which would be higher for an uniform certificate. Coherence All options would improve the domestic and cross-border functioning of the small brewers relief. The increased legal clarity of the regulatory options would increase the ease of doing business for cross-border businesses and ultimately improved their competitiveness. The impact of the non binding options are similar to the regulatory options but the magnitude of their effects may be lower if MS chose not to implement the guidelines.

Option 2: Extending the reduced rates to small cider makers Effectiveness The aim of the existing reduced rates scheme for small brewers is to support the competitiveness of SMEs vis à vis large players. The extension of this scheme to small cider makers would enhanced the competitiveness of these producers with limited adverse effects in terms of foregone revenues and administrative burdens.

Efficiency This option would impact tax revenues for public authorities, however on the whole these impacts are negligible with modest impacts in the traditional cider MS of UK and IE. Small cider makers would gain relatively to large producers but market effects are estimated to remain small, given the limited amount of sales covered by the reduction. In terms of costs for businesses, these would be similar to that of small brewers, which are negligible at EUR 0.32/hl. From an enforcement perspective, due to the numbers of businesses involved, the amount of excise revenues involved and the marginal role of cross-border trade there would be no significant requirement for additional resources. Coherence The public health effects of the reduced rates would be limited with only noticeable impacts in the traditional cider MS, such as IE and the UK.

Option 3: Increasing the threshold for low strength alcohol Effectiveness Low strength alcohol provisions are largely unused due to the low threshold which is irrelevant for most of the beer market, with the exception of radler and a few other beers. The brewing industry has reacted to the health conscious consumer and is developing more low strength beers. It is more costly to brew low strength beers and this relief would support the competitiveness of these products with limited adverse effects in terms of foregone revenues and administrative burdens. Efficiency As this option would reduce the rates for low strength alcohol, it would impact tax revenues for public authorities. These impacts are negligible for MS due to limited volume of sales that would be covered by increasing the threshold at which reduced rates apply for low strength alcohol. There would be no significant, if any, requirement for additional enforcement resources. Similarly there would be little, if any, additional costs for businesses. Coherence It is not clear in the Directive as to the objective of this reduction and therefore it is difficult to judge its coherence. However reduced rates promote an alternative to high strength beers in line with public health objectives.

7.3.2. Stakeholders' views

Respondents to the OPC agreed with the option clarifying the rules for the cross-border recognition of small producers, as well as the rules to determine when a producer is independent. The consensus is almost unanimous within the beer industry, where more than 90% of respondents are in favour of these changes, without significant differences between SMEs and other entities. Also taking into account the whole sample of respondents, more than 60% of them agreed or strongly agreed with this option. The provision of non-binding guidelines while leaving the legislative text unchanged was also positively assessed by respondents, from both the beer industry and the overall sample. However, the support for non- binding guidelines was milder, with about half of the respondents agreeing to this.

The response to the OPC in relation to extending the reduced rates to cider was small and mixed. Producers of OFB (or representative thereof) were somehow more negative than others. However, industry responses should be considered cautiously, as only one respondent out of 31 is exclusively active in the OFB market. Tax authorities either welcomed or did not oppose the possibility of granting reduced rates to small cider makers.

Results from the OPC conducted on the revised threshold for low strength beer also provide a mixed picture. While 47% of participants who responded to this question welcome an increase in the threshold of low-strength beer from 2.8% to 3.5% vol, 44% of participants disagree with this policy option. Most respondents who support the raise in the threshold for low-strength beer are beer producers, while most respondents against it are other alcoholic beverages producers.

7.3.3. Comparison summary and preferred option

2 – Option 1.a.1 – 1.a.2 – non 1.b.1 - small 3 – low No amend binding recognition cider strength change* EFFECTIVENESS Directive guidelines makers ensuring fair treatment of ++ ++ + + + 0 businesses across all alcohol sectors preventing and correcting any ++ ++ ++ ++ + 0 distortions of competition providing clear rules on the scope, classification and calculation of + + ++ + ++ 0 duties for businesses and MS providing clear and efficient conditions to determine n/a denaturation procedures reducing administrative burden and compliance costs for businesses and + + + 0 0 0 tax authorities provide legal certainty ++ + ++ ++ + 0 strengthening the fight against fraud n/a and tax evasion improving human health protection n/a n/a n/a 0 0 0 EFFICIENCY administrative burden + + + 0 0 0 tax revenues 0 0 0 0 0 0 COHERENCE 0 + + + + 0 OVERALL + ++ ++ ++ ++ STAKEHOLDERS OPINION ++ + + + +

Defining the term 'legally and economically independent' is the ultimate aim of option 1.a, which could be achieved by a regulatory or non-regulatory approach. It should be noted that a certain level of consensus already exists among MS authorities and as non-binding interventions have already proved effective in defining the conditions of applying reduced rates to small brewers, therefore the net benefits of using a non-binding instrument would seem to outweigh those of a legislative revision.

In terms of improving the cross-border implementation of this relief, the regulatory ex-ante approach (Option 1.b.1) as a complement to Option 1.a.2 is preferred to the current absence of a harmonised approach, as this would ensure consistency throughout the EU. The preferred option is therefore a combined Option 1.a.2 + 1.b.1.

Option 2 on the extension of the reduced rates to small ciders makers and the increase of the alcoholic threshold (Option 3) to which reduced rates are applicable for beer are the policy choices for their respective problem areas.

The package of options under the cluster of measures relating to reduced rates issues is therefore composed of the combination of Option 1.a.2, 1.b.1, Option 2 and Option 3.

7.4. Unclear provisions to measure Plato degree for sweetened / flavoured beer

7.4.1. Comparison of options

Effectiveness As discussed above, policy option 1 and 2 have the same target (i.e. selecting a harmonised approach for the measurement of Plato degree of sweetened/flavoured beer) but are based on different measures: a regulatory amendment of Art. 3(1) (option 1 and its sub-options) or non- binding guidelines (option 2 and its sub-options). The extent to which the options will meet the policy objectives clearly depends on the degree of adoption / compliance across MS. In the case of option 1 we can assume full compliance by all authorities, while the adoption of guidelines (option 2) would not be mandatory, so MS may not conform to the suggested measurement approach. This distinction is particularly important when it comes to impacts on legal certainty, since the persistence of disparities of interpretation across the EU may eventually encourage rather than decrease the risk of disputes between businesses and tax authorities, especially in MS that would eventually not adopt the Commission’s guidance. As noted above all stakeholders interviewed advised that they would only reluctantly switch away from their current approach unless binding changes are made in the Directive.

Efficiency Policy options 1 and 2 do not pose any (in)efficiency problems although both would require some adjustment costs in the adaptation of some control and monitoring processes. It should be recalled that different MS will have different baseline scenarios, depending on which method they currently use and to which method they would need to switch. From the analysis of impacts it is apparent that selecting approach A or B1 would result in an overall decrease in excise receipts from sweetened/flavoured beer of more than EUR 30 million (about -25%), compared to the baseline situation. Selecting approach B2 would result in relatively smaller changes since this is the approach currently in force in most of MS and also the one generating the highest excise revenues. Moreover, approach B2 that allows authorities to perform checks directly on the end-products, with no need for on-site inspections and/or measurement during the production process, would be more cost-effective. It could be argued that approach B2, as less disruptive and more widespread already, would be, collectively, more efficient.

Coherence Directive 92/83/EC gives MS the choice to levy excise duty on beer on the basis of either the number of hectolitres/degrees Plato or the number of hectolitres/degrees of ABV. The coexistence of the methods was analysed in the Ramboll evaluation, which concluded that this situation created no major difficulties or negative consequences for the internal market. This conclusion was widely supported by MS and beer producers as the Plato measurement is based on long-standing tradition in many MS.

None of the retained options clarifying the measurement method stand in contradiction to this preference and all options are therefore in principle coherent with the legislation and with the smooth functioning of the single market. As stated in the present impact assessment, the problem at stake regarded not the relevance of existence of the Plato/AVB methods but the stakeholders – authorities and businesses - understanding of Art. 3(1) with regard to at what point in the production process the degree Plato should be measured.

Since the choice of option implies switching to one or the other Plato measurement approach, it will inevitably impact some countries more than the others. From the perspective of countries which will need to adjust their processes and procedures to comply with the new approach, the options could be perceived as incoherent in terms of correspondence with the present national

practice. However, given that the lack of coherence in application of the Plato measurement method is the very problem at stake behind the present initiative, the coherence aspects should be seen from the perspective of the single market. In that case all options are coherent with the objective of ensuring smooth functioning of the single market and ensuring coherence of product treatment in each geographical market.

In terms of external coherence, the Plato measurement, being of technical nature, has no perceived impact on other EU policies, regardless of the chosen option.

7.4.2. Stakeholders views

The level of agreement of the OPC participants varies. A small majority of respondents (53%) believe it is necessary to amend Art. 3(1) of the Directive and to clarify the term ‘finished product’ with regard to sweetened/flavoured beer; however, 38% disagree with it. The percentage of stakeholders against an amendment of Art. 3(1) grows if only beer industry respondents are considered (56%, against only 37% in favour of a policy change).

There is instead greater consensus on the need to provide non-binding guidance on this issue: 61% of respondents (and 70% of beer industry stakeholders) are in favour of non-regulatory approach under option 2 and its sub-options. In their qualitative contribution to the OPC, several industry players mentioned the need to adopt either approach A or B1, as approach B2 in their view is ‘technically incorrect’. Interestingly, some respondents have emphasised that the most effective solution would be the application of the ABV method to sweetened/flavoured beer. A few respondents were concerned of the uncertainty and believe that significant room for tax fraud would be generated by selecting approach A or B1.

In some MS included in the studied sample, all stakeholders (including beer producers) would only reluctantly switch away from approach B2. In other MS, brewers exerted some pressure to stop using approach B2, despite the latter being the preferred approach by tax authorities; these countries may be more open for a change. MS currently adopting approach A or B1 are unlikely to change to approach B2 unless binding changes are made in the Directive.

7.4.3. Comparison summary and preferred option

1.A – 2.A – 1.B.1 2.B.1 2.B.2 Option: A (before sugar) 1.B.2 No reg. non – reg. – non – not – reg. B1 (real extract) B2 (present extract) reg. reg. reg. change* EFFECTIVENESS regulatory (reg.) and non-regulatory (non reg.) ensuring fair treatment of businesses across all ++ + ++ + ++ + 0 alcohol sectors preventing and correcting any distortions of ++ + ++ + ++ + 0 competition providing clear rules on the scope, classification and calculation of duties for ++ + ++ + ++ + 0 businesses and MS providing clear and efficient conditions to n/a determine denaturation procedures reducing administrative burden and compliance + + + + + + 0 costs for businesses and tax authorities provide legal certainty ++ + ++ + ++ + 0 strengthening the fight against fraud and tax n/a evasion improving human health protection n/a EFFICIENCY administrative burden + + + + + + 0 tax revenue - - - - 0 0 0

COHERENCE + + + + + + 0 OVERALL - 0 - 0 ++ + STAKEHOLDERS OPINION + ++ + + + ++

The objective of legal clarity in this area is necessary as divergent interpretations of the term 'finished product' exist within the EU. While the regulatory and non-regulatory options would result in similar impacts on the markets, the compliance with these options may differ. Given the clear benefits for all of legal certainty, the options of amending the Directive are the preferred option as being the only one that would ensure compliance. When it comes to choosing between approaches B1 and B2, the key distinction between the two is the efficiency of their implementation. As argued above, the approach B2 is considered – collectively - less disruptive to the internal market as a whole and raising most excise revenues for the MS.

Taking these considerations into account, the preferred option appears to be a legislative revision of the Directive, standardising approach B2 of Plato measurement – Option 1.B.2.

It must be recalled at this point in time that the preferred option stems directly from the objective analysis but it does not take into account the upcoming CJEU ruling. The precise scope and the extent to which the CJEU will clarify all outstanding uncertainties of the Plato situation is unknown. If the CJEU rules contrary to the preference stated above, the former will take precedence and the jurisprudence will be duly reflected in the revised Directive.

7.5. Summary of preferred package of options

This paragraph provides an overview of the preferred options corresponding to the identified problems.

7.5.1. Dysfunctions in the application of exemptions for denatured alcohol

Mutual recognition of CDA

The preferred option is a regulatory amendment of the Directive to ensure that the divergent interpretations involving MS that have notified CDA formulations other than the Eurodenaturant will be eliminated and legal certainty will be achieved (Option 1). This option means a codification of the existing practice. Each MS would have to recognise CDA produced in another MS using the formulations notified by that particular MS, but not those notified by any other MS. This would mean that MS retain control over the CDA produced within their territories, while being obliged to also exempt any CDA legally produced in another MS. This option will reduce any remaining trade barriers and distortions and consolidates MS desires for a harmonised solution for CDA into a legal text.

PDA formulations

The preferred option is a regulatory amendment of the Directive to clarify the unclear wording of the Directive to increase the legal certainty for indirect uses and ‘finished product’ containing PDA (Option 4) accompanied on an optional basis by Option 3 (capacity/confidence building measures). The clarification would make reference to a 'finished product' in order to provide MS with flexibility for the various product groups using PDA. This option would also define a quantitative line above which a product containing denatured alcohol must always be moved in accordance with Chapter IV of Directive 2008/118/EC. This will be included as an amendment to the Directive. This option would eliminate the ambiguity and uncertainty that currently exists in relation to PDA. Moreover, it would ensure equal treatment of goods containing PDA across

the EU and reduce the risk of costs associated with disputes between businesses and national authorities.

7.5.2. Dysfunctions in the classification of certain alcoholic beverages

The preferred option is to split the OFB category into two subcategories of which one would maintain the current treatment, while the other would ideally comprise of all traditional OFB products (i.e. cider and perry etc.) which would be defined and treated separately (Option 2.a). This option aims to differentiate the OFB products that arguably correspond to the original definition and intention of the legislator from the ‘novel’ products that have been opportunistically designed to fit into it or simply that do not fit elsewhere. This option would result in legal certainty at EU level and a consistent treatment of borderline products across MS since it would make the current national level non-harmonised distinctions unnecessary. This option can be complemented, on an optional basis, by Option 3.a /Option 3.b/Option 4.a/Option 4.b.

7.5.3. Dysfunctional application of reduced rates

Legally and economically independent small brewer

The preferred option is to define the term ‘legally and economically independent’ by non- binding guidelines (Option 1.a.2). Such definition would encompass the general norms and principles as well as detailed technical specification outlining the legal conditions which could determine if companies are independent or not. There is already a certain level of consensus among MS authorities and non-binding interventions have already proved effective in defining the conditions of applying reduced rates to small brewers.

To ensure that the conditions for recognition of small brewers are the same in each MS, the preferred option is to identify small brewers through a uniform certificate, defined via a Commission Implementing Regulation (Option 1.b.1). This certificate would need to be presented when a small brewery would like to claim reduced rates in a MS other than that of establishment. This certificate should be provided, upon request, by all customs authorities to all businesses up to 200,000 hl, regardless of whether they can access reduced rates in their country of establishment. This certificate could be developed through the Fiscalis programme and would guarantee equal conditions for small brewers active across borders.

Extending reduced rates for small producers to other sectors

To address the unfair competition between small producers of alcoholic beverages, the preferred option is to amend the Directive and extend the reduced rates to small cider makers (Option 2). As for the small brewers reduced rates, this reduced rate would remain optional for MS. It would be based on the definition of an independent producer and a maximum discount rate compared to the standard rate would be fixed. The maximum yearly output threshold would be 15 000 hectolitres per year to allow small cider makers to benefit from the reduced rates if MS make use of the option to apply a reduced rate. This option has limited impacts in terms of costs and would improve the competitiveness of cider makers.

Increasing the threshold for low strength beer

The preferred option is to increase the threshold to which reduced rates are applicable to beer as this would encourage the development of low strength beers (Option 3).

7.5.4. Unclear provisions to measure Plato degree for sweetened / flavoured beer

The preferred option is to clarify the definition of ‘finished product’ by outlining when the measurement of Plato degree should occur (Art. 3(1) of the Directive) (Option 1.b.2). This option consists of a regulatory amendment to clarify that the term ‘finished product’ refers to the end-product that is released for consumption, meaning that that sugars or flavours added after fermentation would contribute to the Plato degree. This option would provide legal clarity of the term ‘finished product’. The regulatory amendment will ensure full compliance and is the least disruptive of the internal market, taking account of the current approaches on national level.

8. REFIT (SIMPLIFICATION AND IMPROVED EFFICIENCY)

8.1. Context, methodology and constraints

Revision of Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages is part of the Commission's REFIT programme. One of the original objectives behind the Ramboll study was to identify weaknesses in the legislative environment caused by the Directive resulting in negative consequences for the stakeholders (e.g. obstacles to the functioning of the internal market, competitive disruptions, administrative and compliance costs.)

Before analysing further, it is important to understand that despite this original level of ambition, the Ramboll evaluation28 and the Study clearly concluded that, overall, Directive 92/83/EEC did not directly impose compliance costs on economic operators. Instead by including certain products in the scope of excise duty, the Directive indirectly subjected those products to the provisions of Directive 2008/118/EC, which sets out the rules and conditions for holding and moving excise goods. Additionally, MS exercise some level of flexibility regarding provisions at national level and requirements regarding certain procedures (see below).

The resulting room for diverging interpretation since 1992 involuntarily allowed economic operators, as well as national tax administrations, to establish their own modus operandi. Most of the problems identified – as described in the problem definition section – were specific to certain markets or certain products. In terms of REFIT objectives, these focused particularly on those areas where economic operators reported burdens. Taking into account the considerations presented above, in the case of Directive 92/83/EEC the REFIT discussion is therefore shifted from not so much the excessive costs and burdens to unnecessary costs and burdens, which could be avoided if the Directive functioned better.

Overall, evidence collected in the Ramboll evaluation, the Study and feedback gathered from the day-to-day application of the Directive's provisions led to the conclusion that there was nevertheless a perceivable – albeit hardly quantifiable – lack of legal certainty over the treatment of specific products, leading in turn to potential additional costs to economic operators. The lack of certainty could be classified under the 'hassle' or 'irritation' costs, which are often linked to administrative burdens and constitute residual category of direct costs, which are difficult to quantify or monetise and to relate to a specific information obligation. Such costs could include administrative delays, opportunity costs of waiting time, etc. The stakeholders were not in a position to provide any estimates of the monetary impacts of the lack of legal

28 See: Chapter 2.5, p. 36 of the Ramboll evaluation

certainty; what we have obtained were the subjective opinions of the best placed stakeholders: the economic operators and administrations. These aspects are nevertheless an important indicator of the 'well-being' of the stakeholders.

As stated above, the majority of problems relate to the legal uncertainty that the economic operators experience with production, use and/or movement of some alcoholic products (governed by different law even though the stakeholder may not be aware of it). For example, in the area of denatured alcohol, the main concern of the economic operators regarding administrative burdens was linked to the specific requirements regarding supervision of production and movement of products containing denatured alcohol, which cannot be directly linked to the provisions of the Directive 92/83/EEC, and which represented a mix of compliance with the above-mentioned Directive 2008/118/EC or the national-level response of some MS to their estimations of the risk of fraud. In the area of classification, the Ramboll evaluation concluded that the classification of most alcoholic beverages from an excise perspective was generally straightforward and resulted in little to no direct administrative burdens. It identified at the same time costs resulting from the complications and disputes arising from situations in which the stakeholders disagree on the correct interpretation of the provisions of the Directive.

Therefore, in the context of the present initiative, the REFIT aspects related predominantly to identifying opportunities for simplification, reduction of inconsistencies, gaps and other ineffective measures which can lead to unnecessary costs. Most of the opportunities are linked to elimination of the legal uncertainty over the interpretation of certain ambiguous provisions.

Both the Ramboll Evaluation and the Study attempted to gather estimates of these costs. Unfortunately, only anecdotal evidence (and without monetised disadvantages) was available where the problems resulted in legal disputes before the CJEU. For example, in terms of the classification problems, the economic operators were not in a position to provide precise monetary quantification of the expected cost due to the varied nature of the legal cases reported (e.g. depending on the evolution of a given case, the economic importance of the disputes, the willingness of the parties to settle the matter via the judicial system, etc.). Some anecdotal evidence was provided by a few MS or economic operators, relating to specific cases. Such evidence is duly reported under the problem definition of this report to illustrate the problems, but cannot stand for the baseline against which any cost and burden reduction measurement could be calculated.

Having no baseline, it was equally, if not more difficult to estimate any potential benefits of the proposed changes. This difficulty is reflected in the table below, where the analysis of the expected regulatory benefits is presented qualitatively. Any estimates provided are often hypothetical, based on a rigorous set of assumptions which were explained under each specific option under the analysis of impacts. Moreover, most of the quantification relate to the cost side of the REFIT given that most of the benefits did not have a quantifiable base to start from. That should by no means indicate that there would be no REFIT-type benefits stemming from the initiative. To the contrary, the Study concluded that the additional regulatory costs to comply with any new rules are mostly one-off and not significant in the broader scale, quickly offset by the benefits. The difficulty lays in the lack of numerical baseline values for most of the data.

To conclude, it should also be noted that the burdens stemming from (mostly) legal uncertainty would have been burdens only to the businesses operating fairly in the markets. The burdens for them would however be an opportunity for those businesses who intended to profit from the unclear legislation by, for example, marketing products that would resemble high alcohol content products taxed at a higher rate but which would fall under the preferential OFB category. In such situations, the net beneficiaries of the initiative would be the honest businesses trying to comply while being exposed to unfair treatment. Since the Study concluded that the

additional costs and burdens for any solution were found marginal and off-set by benefits, it could be concluded that the net outcomes will be globally positive for all stakeholder negatively affected by the status quo.

Summary of REFIT costs and costs reduction

REFIT Cost reduction – Preferred Option(s) Main Description Estimates Comments beneficiaries Minor positive impacts for producers that sell CDA to The reduction of the hassle costs – MS with different national and subsequently elimination CDA producers formulations, and users of thereof - associated with the n/a operating cross- CDA in these MS stemming disputes and delays due to non- border from lower risks of disputes recognition of CDA methods

with authorities of the were not possible to estimate

receiving MS

Cost savings stemming from The savings stemming from the enhanced clarity surrounding legal costs related to disputes over PDA users and the legal meaning and uses of the PDA and their use in other producers operating PDA which would ensure products - and subsequently

PROBLEM 1 cross-border equal treatment of goods elimination thereof- are case-

Denatured alcohol Denatured containing PDA across the specific and the baseline values n/a National EU and reduce the risk of were not reported by the administrations costs associated with disputes stakeholders to allow for (customs between businesses and estimations of benefits laboratories) national authorities

Overall burden not expected to change significantly (€ 2.0 – Legal certainty at EU level 2.5 million)

and consistent treatment of Cider/perry One-off reclassification costs of borderline products across Familiarisation cost: producers across the familiarisation costs, updating of MS. However, the distinction approx. €4,500 per EU the IT systems, and national between the products may company (including procedures, training for economic lead to the creation of new overheads) or National operators are to be expected. borderline products which aggregated burden of administrations These costs would be offset by could, in worst case scenario, € 4.5 million

Classification PROBLEM 2 the benefits in 10 years or more neutralise benefits of any new clear definition IT updates: approx. €800 per company or aggregated burden of to € 6.9 million.

Recognised small brewers: total burdens for 675 operators in the sample MS: approx. In terms of recognising the €13 000 or 2% of the

Established / recognised small status of a legally and burdens estimated for brewers would incur limited economically independent the overall scheme; administrative or enforcement brewery, more legal clarity costs (equalling to asking for the Small breweries and ease of doing business for Not yet recognised certificate), while these who are across the EU cross-border economic small brewers: total not recognised as small brewers operators will result thanks to burden for 180 would need to prove their status

PROBLEM 3 Reduced rates Reduced the EU-wide certificate for operators in the first small breweries sample MS not under the scheme: approx. €32 000 or 5% of the burdens estimated for the overall scheme.

Annual burdens per In terms of extending the small cider maker: reduced rate scheme to small approx. €178 per cider makers, the burdens Small cider economic operator or associated with compliance producers across the an aggregated total with the scheme would be EU for the sector of similar to those incurred by €200,000 annually small breweries

Legal certainty and reduction The amount of legal costs related in legal costs of judiciary to disputes over the measurement disputes stemming from method for excise tax base were Breweries eliminating disparities of n/a not provided by the stakeholders, National interpretation of Plato which makes it impossible to administrations measurement methods estimate savings linked to their

Plato degree PROBLEM 3 across the EU elimination

9. HOW WOULD ACTUAL IMPACTS BE MONITORED AND EVALUATED?

The monitoring of the implementation and functioning of the revised rules will be role of the ExComm, an advisory committee on excise issues chaired by the Commission in which representatives of all MS participate. The ExComm will report on any problems with the implementation and the evolution of problems with the functioning of the Directive as addressed in this impact assessment, and discuss and clarify possible interpretation issues between MS regarding the new legislation. In case new legislative developments are required, the ITEG might be further consulted.

MS and the Commission will evaluate the functioning of the evolutions provided for in the new legislation. To that purpose, MS will communicate to the Commission any relevant information as regards the level and the evolution of the regulatory costs, legal certainty, economic distortions and market abuse, excise fraud, etc. necessary for the evaluation of the effectiveness, efficiency, coherence with other interventions with similar objectives, and continued relevance and EU added value of the new legislation. The evaluation should also seek to collect input from all relevant stakeholders as regards the level and the evolution of their administrative burden and compliance costs or instances of market distortions. The Commission will prepare the evaluation at the earliest 5 years after its entry into force, allowing the markets to adjust and the results and impacts to materialise.

Without prejudging the exact scope and extent of the future evaluation and the ongoing monitoring, both of which will live and evolve together with the functioning of the revised Directive, the tables in Annex 18 provide an indicative overview of key expected results and/or impacts and accompanied by examples of possible indicators expected to feed into their assessment.

The indicators are set either at the result-level (e.g. number of instances of non-compliance, number of law cases, existence and number of diverging interpretations, reduced cross-country disparities, etc.) or at the impact-level (e.g. changes in the market structure of the OFB, revenues from excise duties, improved competitiveness, reduced scope for misclassification, costs savings and investment, etc.). The result-level indicators can and will be regularly reviewed through the works of the committees and the Commission and will feed into the future evaluation. The impact-level indicators, given their far-reaching nature, sheer complexity and burdens associated with their collection and/or assessment, will only be analysed at the moment of the retrospective evaluation through a multi-pronged approach involving many stakeholders and detailed data. This distinction is marked in the monitoring and evaluation table in Annex 18 and is important to retain.

Additionally, since the industry producing and/or using alcohol and alcoholic beverages is active and closely follows the work of the Commission, it is expected that any issues related to the application of the new rules, would be reported without much delay directly by the stakeholders. That could be done either by contacting the respective Commission services or through tabling of motions for actions through the REFIT Platform, for example29.

It would have been preferable to set success criteria and benchmark values for the expected changes. However, having no firm value for most of the problems (detailed analysis and explanations are included in the annexes relating to drivers of the problems as well as Chapter 8 on REFIT considerations), it is unfeasible to set measurable targets. Presently, it is only possible to foresee analysis of trends or market structures, which would be done mostly through a full economic study accompanying the future evaluation.

29 Proposals tabled to the REFIT platform already took place in the excise duty on alcoholic beverages: XVIII.12.b on reducing the room for diverging interpretations of rules for the wine and spirits industry and XVIII.12.a from the whiskey producers;

10. ANNEXES

ANNEX 1. PROCEDURAL INFORMATION

1. Lead DG, Decide Planning/CWP references

The lead DG is DG TAXUD.

This initiative got the following political agreements: - Agenda Planning: Proposal for revision of Council Directive 92/83/EEC of 19 October 1992 on the structures of excise duty applied to alcohol and alcoholic beverages. (2017/TAXUD/005) - Inception Impact Assessment: Proposal for revision of Council Directive 92/83/EEC of 19 October 1992 on the structures of excise duty applied to alcohol and alcoholic beverages. (Ares(2017)1097709) - Commission Work Programme: 2017 Annex II initiative 7

2. Organisation and timing

The following DG were invited to the Inter-Service Steering Group (ISSG): AGRI, COMP, JRC, GROW, OLAF, RTD, SANTE, SG, SJ, TRADE.

An independent study was carried out in 2014/2016 by a consortium led by Ramboll Management Consulting.30 The recommendations and findings of the Ramboll Evaluation were taken into account in the Commission’s report submitted to the Council in October 2016 (hereinafter the ‘Commission Report’)31. A grouping led by Economisti Associati s.r.l. and including the Centre for European Policy Studies (CEPS), CASE - Center for Social and Economic Research, wedoIT-solutions GmbH, and ECOPA (hereinafter collectively referred to as “the Consultant”) undertook the assignment titled “Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages” (“the Study”). The objectives of this Study were to gather and analyse the evidence on the existing costs and benefits arising from the Directive, with the main focus on analysing the scale of the problems identified in the Ramboll evaluation. The Study also assess the evolution of the problems if no further action at EU level is taken (dynamic baseline scenario) and the economic, social and environmental impacts of the possible options to address the problems identified. Additionally, the Consultant assisted the Commission in conducting an Open Public Consultation (OPC) to collect stakeholders’ comments and feedback on the issues identified and the possible options for a revision of the Directive. The ISSG reviewed and approved the following documents - all ISSG meetings' minutes - questionnaires (OPC, Member State, economic operator) for the stakeholders' consultation

30 Ramboll Management Consulting, Coffey, Europe Economics, “Evaluation of Council Directive 92/83/EEC on the harmonisation of the structures of excise duties on alcohol and alcoholic beverages”, 2016 31 'Report from the Commission to the Council on the evaluation of Council Directive 92/83/EEC on the structures of excise duties on alcohol and alcoholic beverage', Brussels, 28.10.2016, COM (2016) 676 final.

- inception report32 of the independent contractor's study - final report of the independent contractor's study - Commission's Impact Assessment Report (this document)

The chronology of the main events is as follow: 26/01/2017 first ISSG meeting 29/03/2017 ISSG meeting on the inception report April 2017 approval of the open public consultation (OPC) questionaire Apr – Jul 2017 stakeholders' consultation, including OPC in all EU languages 04/07/2017 Presentation of the Progress Report to the ISSG Jul - Aug ISSG review of the study's draft final report 07/09/2017 ISSG meeting on the study's draft final report 06/11/2017, 12/12/2017 ISSG meeting on Commission's Impact Assessment Report

3. Consultation of the RSB

The Impact Assessment report was examined by the Regulatory Scrutiny Board on 24/01/2018.

Overview table of the changes compared to the first version submitted to the Regulatory Scrutiny Board

Board recommendation What has been done? Where?

1. The report does not adequately Further information on the Section 1 frame this initiative in the context of REFIT evaluation of Directive the evaluation and the positions 92/83/EEC has been included expressed by different stakeholders. throughout the report in order to address this recommendation. Clarification on the scope of the initiative has been added to the introduction and content section.

The positions expressed by the Sections 1, 2.1 different stakeholders have been and Annex 2 further addressed in the opening (synopsis report – sections and the revised Annex stakeholder 2, which provides the synopsis consultation) report on the stakeholder consultation.

2. The report does not systematically A dedicated REFIT section in Section 8 assess the overall effects of the the report has been included to proposed REFIT measures in terms systematically analyse the of simplification and reduction of effects of the options. administrative burden.

32 Intermediate deliverable, describing the problems, options, methodology and approach of the study

3. The composition and impacts of Further clarification regarding Section 2 the different options are not the complexity of the issues at adequately presented. stake has been included in section 2. Section 5.1 An overview of the drivers, problems, objectives and options has been included in the report in order to provide greater clarity. This outlines the distinct problems and the need to address these in different ways. Section 5.2, 5.3, Further clarification in the 5.5 option section has been added to address this concern.

Further information on the Section 6 impacts of the different options has been included in the report.

4. A summary presentation of the An additional subsection Section 7.5 package of measures considered as providing a summary the preferred option is missing and it presentation of the packages of is unclear why alternative packages measures considered as the are not considered. preferred option has been included in the report. Section 5 Further clarification has been added to the section on options to address the fact that due to the complexity of the issues at stake, alternative options are in some cases not possible.

Further considerations and adjustment requirements of the Board

Board recommendation What has been done? Where?

(1) Justification of limited scope Clarification on the scope of the Section 1 The context should more initiative has been added to the comprehensively build on the findings introduction and content section. of the evaluation and show how the report addresses its recommendations. The report is not part of an Section 1 The report should indicate why the exhaustive revision of excise proposed initiative is not part of a more duties on alcohol. This is due to exhaustive revision of excise duties on the limited stakeholder support alcohol (e.g. regarding excise duty for amending minimum rates.

rates). It should summarise This clarification has been stakeholders' views and expectations included in section 1 of the (e.g. through references to Council report. conclusions or REFIT opinions). Section 1 The report should explain how this A revised synopsis report on the Annex 2 initiative will be complemented or stakeholder consultation is Annex 5 modified by future health policy. included in Annex 2. All of the above should deliver a Clarifications on stakeholders' rationale for the limited scope of this views and expectations have also initiative and a better correspondence been included in the report. The between the problem definition, Council Conclusions have been stakeholders' concerns, stated added to the annexes. objectives (e.g. on health) and the Section 2.3, 2.4, proposed response. Additional information has been 2.5, Annex 9A added to the report to address the Section 5 recommendations of the REFIT evaluation of the Directive. Section 6 The report does not include any options that endanger human health and where appropriate we include public health in the section on the impacts of options. Public health impacts are relevant only in the problem of denatured alcohol and low strength alcohol. The modest relevance of the options for public health is inherent in the type of reform envisaged.

(2) REFIT section A dedicated REFIT section in Section 8 A dedicated REFIT section in the report the report has been included to should systematically analyse areas systematically analyse the where the proposed measures introduce effects of the options. new obligations (e.g. new certificates for small producers) and those where it This section outlines the benefits lifts some existing requirements. of this proposals in terms of This section should provide a clear reducing unnecessary costs and conclusion on the overall net reduction burdens, if the Directive was (or creation) of administrative burden functioning better. Stakeholders stemming from the proposed package were not in a position to provide of measures and should describe who it estimates of the monetary will affect. It should justify the impacts of adminstrative introduction of new requirements in the burdens, however their light of objectives other than subjective options and anecdotal simplification (e.g. to enhance clarity evidence was obtained. and legal certainty). Finally, this section should also better The section includes a summary

delineate the scope of administrative table of the expected regulatory burden that falls in the remit of the benefits of each option. proposed initiative and distinguish it from obligations that emanate from The section also refers to other pieces of EU, national or local obligations from Directive regulations. 2008/118/EC, which sets out the rulesand conditions for holding and moving excise goods, including alcohol and alcoholic beverages.

(3) Packaging of proposed measures An additional subsection Section 7.5 The report should clarify the packaging providing a summary of the proposed measures. It should presentation of the packages of better highlight choices to be made measures considered as the between different measures and preferred option has been distinguish areas where there is no included in the report. alternative option. The report should Section 5 also better explain how individual The report considers the measures combine and complement preferred options to be those that each other. can achieve the necessary impact as a standalone measure. Complimentary options which do not work in isolation are not included as the standalone preferred options.

Further clarification has been added to the section on options to address the fact that due to the complexity of the issues at stake, alternative options are in some cases not possible.

(4) Impact of options An overview of the intervention Section 5 The report should more systematically logic is included in the report to take into account the different illustrate the problems, the objectives of the initiative when drivers, the objectives and the assessing the impacts and effectiveness corresponding options better. of the different options. It should better These distinct problems and describe how each measure tackles their underlying drivers need to different problems and contributes to be addressed in different ways the achievement of different and this is reflected preferred (sometimes possibly conflicting) options. Section 6 objectives. Further information on the impacts of the different options has been included in the report.

Section 7.2 Preventing circumvention of It should also include a presentation of legislation is difficult to achieve,

potential risks and trade-offs (e.g. however the new rules for between simplification and legal distinguishing cider and perry certainty). For instance, reviewing the within the other fermented classification of alcoholic beverages beverages category is preferred. appears to aim both at tackling abuses This option would avoid of the system by business operators and instituting differences in the tax at providing enhanced clarity to tax category which may authorities. The report should better unintentionally exclude some differentiate the performance of the eligible products, erode this tax proposed measures vis-à-vis these two category in legal terms and objectives and should convincingly increase the overall complexity demonstrate why a new classification of the system. would actually prevent business Section 7 operators from circumventing the new rules. The scoring system has been revised inline and reflects the The scoring system for the comparison detailed comments and choice of of option should be revised and preferred options. complemented to make sure that the assessment matches the more detailed comments and choice of preferred measures. (5) Comparison of options The tables displaying the Section 7 Since readers may place different levels comparison of options have been of importance on various pros and cons amended to take account of this of the options, the comparison of recommendation. The ratings are options should not add individual now + / - or 0 if there are no ratings and report total scores as a pros/cons for the option. There is single figure. no total score for any option.

(6) Summary of preferred options An additional subsection Section 7.5 The report should include a summary providing a summary description of the preferred option, as presentation of the packages of well as a section grouping the ones that measures considered as the were discarded with explanations why preferred option has been each was dropped. included in the report. Section 5.2, 5.3 Discarded options have been included under the cluster of options dealing with a specific problem. The presentation of the preferred option Section 7.1, 7.2, should clearly indicate the choices Clarification has been added to 7.3, 7.4 made that result in this specific the section on how the options combination of proposed measures compared clearly indicating the forming a coherent whole. In particular, preferred options and the the report should clarify what were the reasoning for this. determining factors to select this preferred option. Section 7.5 It should explain the implementation The additional subsection requirements and describe what is outlining the summary of the likely to happen once all measures are preferred packages describes

implemented. what is likely to happen once all measures are implemented.

ANNEX 2. SYNOPSIS REPORT – STAKEHOLDER CONSULTATION

1. Consultation strategy

Three different types of consultation activities took place, each of them tailored to the targeted type of stakeholder:

1. Consultation of the Indirect Tax Expert Group and Fiscalis Project Group

2. Targeted questionnaires and in-depth-interviews of key stakeholders

3. Open Public Consultation

Directive 92/83/EEC has direct effect on public authorities in Member States, economic operators and their related organisations/associations and NGOs active in the health area. The stakeholder groups directly affected by the Directive were consulted by targeted questionnaires, followed up by telephone and in-depth interviews as appropriate. Moreover, for these groups round table meetings were organised in 2017 covering the same topics as in the in-depth interviews. All stakeholders, both directly and indirectly affected, including EU citizen, were consulted by the Open Public Consultation. Due to the more indirect relationship of the Directive and the fact the Directive allows Member States flexibility in implementation, the interest of the general public is often low. The influence of the general public on the Directive is also quite low.

The aim of this in-depth consultation programme was to get a better understanding of the overall functioning of the Directive, the necessity of an intervention and the details of the issues at stake. In addition, the objective of the consultation activities was to gather the views of the main stakeholders on a set of possible options for the revision of the Directive.

2. Indirect Tax Expert Group and Fiscalis Project Group

Member States have been consulted through the Indirect Tax Expert Group and in particular on denatured alcohol in the Fiscalis Project Group.

The Fiscalis Project Group (FPG) was created to explore the possibility of applying common denaturing procedures for CDA. Between 2009 and 2016, the Fiscalis Project Group on the Eurodenaturant (phases 1 and 2) met a total of 25 days (including one 3-day seminar), with an average of 25 MS and 3 Commission participants, which equates to a total of at least 700 days of staff time. After the first phase of the project, the FPG proposed a formulation commonly referred to as the '3-3-1' Eurodenaturant formulation, which was adopted on 1 July 2013 by Commission Regulation (EU) No. 162/2013. However, most MS chose to recognise this formulation in addition to, rather than instead of, their national formulations. Furthermore the uptake of the '3-3-1' Eurodenaturant formulation was limited as many economic operators indicated that it was too costly to produce and the availability of one of the elements was problematic. In response to these criticisms, a revised formulation was developed in the second phase of the FPG. This formulation, commonly referred to as the '1-1-1' Eurodenaturant formulation was adopted in 2017 and is currently used by 25 Member States.

Exploratory work on a harmonised list of PDA formulations has also been carried out within the Fiscalis Project Group for certain sectors, namely (1) perfumes, cosmetics and (personal) hygiene products, and (2) screen wash, de-icer and anti-freeze. The discussions regarding the former turned out to be particularly difficult, mainly because the national approaches are

currently so different, with some MS authorising denaturation with ingredients of the final product (such as essential oils) in specific cases, while others have a strictly defined list (or a single formulation) that applies to all producers equally.

3. Targeted questionnaires Overall, 161 stakeholders were consulted, for an estimated total of over 215 individual participants (many interviews were attended by multiple participants). The external contractor conducted interviews in the six countries selected for core fieldwork (DE, FR, IT, PL, RO, UK), as well as in the other six MS selected for the thematic research on specific issues (AT, BE, CZ, ES, FI, NL). The geographical distribution of interviews is provided in Table 1 below.

Table 1 – Breakdown of in-depth interviews, by type of respondents and country of origin

Respondent Type No. of Country of origin No. of stakeholders stakeholders interviewed interviewed European Commission staff 5 France 26 MS Competent Authorities 33 Germany 15 - Tax/customs authorities 21 Italy 19 - Public Health authorities 7 Poland 17 - Other (Agriculture authorities, etc.) 5 Romania 10 Economic operators and trade associations 112* United Kingdom 14 - Beer sector 30 Austria 6 - Wine sector 34 Belgium 8 - Cider and OFB sector 33 Czech Republic 6 - Spirits and liqueurs sector 30 Finland 9 - Industrial alcohol sector 32 Netherlands 9 - Other (e.g. home brewers association, etc.) 5 Spain 4 Public health NGOs 7 Other MS** 2 Others (e.g. experts etc.) 4 EU level 16

Grand total 161 Grand total 161 Notes: (*) the total for this category does not add up to the number of interviews per sector, since various interviewees operate in more than one sector; (**) from MS not included in the selected sample.

All interviews were based on the checklists for discussion that were developed in the inception phase of the external study and further refined and consolidated during the data collection phase. The results of the targeted consultations are discussed in detail below.

3.1. Results of the in-depth-interviews Classification of alcoholic beverages  In all interviews with stakeholders it was confirmed that ‘borderline’ products are generally found in the ‘value’ segment of the market, where tax optimisation is more important than for premium brands.  Nearly all the officers interviewed on this point affirmed the uncertainty with classification of other fermented beverages due to the current definitions and the fact that the distinction between CN 2206 and CN 2208 remains unclear still.  The customs administrations interviewed were generally not in the position to estimate the frequency of problematic products cases, and the administrative burden attributable to these dossiers,  A significant share of stakeholders met during the fieldwork, tax authorities in particular, would be in favour of clearer common criteria for the identification of products that have lost their essential fermented character, than those laid down in CNEN note 2206 00, which reportedly leave the margin for subjective interpretation too ample.

Reduced rates  In terms of administrative burden for economic operators and enforcement costs for public authorities, the fieldwork confirmed that the reduced rates for small brewers do not require unnecessary efforts, from neither companies nor customs. Enforcement costs with respect

to domestic producers were considered to be minimal by all tax and customs authorities interviewed.  Economic operators consider that the provision supports the competitiveness of small brewers, and that the tax discount does benefit small operators, rather than being passed through the value chain.  Very few tax authorities expressed support for the extension of reduced rates to small wine producers, mentioning that they already enjoy significant advantages, such as the zero rate and the simplifications provided by Directive 2008/118/EC.  Tax authorities either welcomed or did not oppose the possibility of granting reduced rates to small cider makers.

Excise duty exemptions for denatured alcohol  The majority of stakeholders consulted (including both national authorities and economic operators felt the current rules at EU level, although complex, were fit for purpose, and there is no need for any fundamental changes to the current framework. Nonetheless, problems can and do occur due to (1) an incomplete / inconsistent mutual recognition of CDA, (2) the proliferation of national regulatory approaches to PDA, and (3) divergent interpretations of certain terms related to PDA.  The interviews conducted as part of the study confirmed that the vast majority of both national authorities and economic operators agreed that there are good reasons to allow MS to define their own rules for PDA.  The interviews with economic operators conducted revealed a mixed picture as regards cross-border trade in PDA. While most interviewees acknowledged the fact that the different procedures and regimes in each MS can and do make cross-border trade more difficult, none of the interviewees felt this had affected them in a significant way, or were able to point to instances where they had incurred unforeseen costs.

Calculation of excise duties on sweetened or flavoured beer using the Plato method

 All stakeholders interviewed confirmed that the EU market for sweetened/flavoured beer is relatively small, but growing.  MS currently using the Plato method to calculate the excise duty having adopted an approach (not) taking into account the added sweeteners and flavours after fermentation are unlikely to change approach unless binding changes are made in the Directive.

4. Overview of the results of the Open Public Consultation (OPC)

The OPC was carried out in the framework of the Assignment.33 The English version of the OPC was launched on 18 April 2017, followed three weeks later by the other versions translated into all the EU official languages. It remained open until 11 July 2017, for a total of 12 weeks. a) Questions The OPC questionnaire consisted of 58 questions, divided into six sections, including one introductory section about the respondent’s profile, four thematic sections, and a final section for the upload of additional documents.

33 https://ec.europa.eu/taxation_customs/consultations-get-involved/customs-consultations/public-consultation-structures- excise-duties-applied-alcohol-and-alcoholic-beverages_en

To take the disparity of respondents’ background into account, each thematic section included general questions suitable for all type of respondents, and more specific questions requiring a more in-depth knowledge of (or specific interest in) the technical functioning of Directive 92/83. b) Profile of the participants

 The OPC attracted a total of 16634 responses; a somewhat low number possibly due to the fact that many companies submitted a joint reply to the OPC via their EU-level industry associations, as emerged during the fieldwork of the Study. Respondents from 21 EU MS participated in the OPC. France and Austria contributed relatively greatly, with 30 and 26 respondents respectively. Other significant countries in terms of absolute number of contributions were the United Kingdom, Italy and Poland. Moreover, 21 responses were collected from EU-level or multinational entities.35  The majority of respondents were economic operators (62) and industry associations (56) and other similar entities.36 In addition, 37 private individuals took part in the survey, while the participation of the other respondent groups was more limited: only 2 public health NGOs, 3 public authorities and 6 miscellaneous respondents.37 For this reason, throughout the rest of this annex, these respondents have been grouped together into a single “other” category.  The majority of industry respondents (including both economic operators and industry associations and other similar entities) were brewers (40 out of 118, i.e. over one third), followed by those involved in the production of other fermented beverages. The other product categories (wine, intermediate products, ethyl alcohol and industrial alcohol) were also fairly represented, with the participation of at least 20 stakeholders per area of activity (see Table 2).  As far as the size of economic operators is concerned, the majority of respondents were SMEs (including micro, small and medium size companies, for a total of 34). In addition, 22 large companies with more than 250 employees also took part in the questionnaire.

Table 2: Overview of respondents to the 2017 OPC – breakdown by category

34 The total number of responses initially reached 171. However, after a preliminary analysis three entirely blank responses and two duplicates were discarded. 35 A number of EU-level organisations had improperly classified themselves as being based in Belgium and were therefore reclassified. 36 Note that a number of companies had erroneously classified themselves as “industry associations” and therefore had to be reclassified as “economic operators”. Moreover, certain entities – self-declared as NGOs but having specific interests in certain products and sectors – were included in the “industry associations” group, which was for this reason renamed as “industry associations and other similar entities”. 37 The 6 miscellaneous respondents include a trade association of farmers, an advocacy firm, an alembics manufacturer, a canning company, a private consultant for farmers and a consulting company for the beverage industry.

45 40 40 33 35 31 30 25 26 25 23 20 20 15 10 5 0 B W OFB Int Eth IA Oth

Legend: B: production of beer; W: production of wine; OFB: production of other fermented beverages; Int: production of intermediate products; Eth: production of ethyl alcohol; IA: production or end-use of alcohol for industrial uses; Oth: other (e.g. production of fermentable raw materials, distribution and retail of alcoholic beverages).

Note: the sum of respondents by area of activity exceeds the total number of economic operators/industry associations since many of them operate in more than one area of activity. c) OPC results per thematic section

The section that attracted the greatest interest was that on the reduced rates or exemptions for certain producers and types of alcoholic beverages, which totalled a number of responses ranging from 68 to 132. Table 3 below provides a more detailed overview of the number of responses per thematic section

Table 3: Overview of respondents to the 2017 OPC – Number of respondents by sub-section (not including “Don’t know” answers)

Reduced rates or Denatured Classification Plato method exemptions alcohol

Minimum and maximum 51 - 131 68 - 132 16 - 82 50 - 60 number of respondents

Note: The number of respondents across different questions within the same section varies, therefore the range between the question with the lowest number of responses and that with the highest number is indicated.

Section 1: Classification of alcoholic beverages

All industry stakeholders tend to agree that there is no need to reconsider the tax treatment of RTDs, beer-mixes, fermented-base liqueurs and high strength fermented beverages, with the only exception of beer producers who – while they see no issue with beer-mixes – consider that the treatment of the other products may require a partial revision. To the contrary, the majority of private individuals and of respondents falling into the residual ‘other’ category believe the tax treatment of the products, and especially RTDs, needs to be reconsidered.

A clear majority of the industry opposes a different treatment of new mixed drinks from more traditional alcoholic beverages, whereas private individuals tend to agree with the principle. Similarly, the industry opposes special taxes on products intended for the youth, while private individuals and the ‘other’ respondent category are in favour. Unsurprisingly, the taxation of products that are equivalent for consumers but based on fermented or distilled alcohol divides the industry, with producers of fermented beverages (beer, wine and OFB) in favour of different levels of taxation depending on the base (fermented or distilled), and spirits producers advocating for an equal treatment, regardless of the base. To the contrary, all respondent groups

agree that wines and beers using alcohol as flavour-carrier should not be taxed more heavily for this.

While the above clearly reflects the position of the industry, preferring the status quo and expecting negative impacts of a revision of the classification of the definitions of alcohol and alcoholic beverages, the actual impact assessment shows positive impacts of a revision of this area. The position of industry in this specific section is in line with the overall position of the industry – preference for the status quo because the results of the revision are unknown for this group, especially after having the same legislation for the last 26 years. In addition, it is clear when the respondents are distinguished by interest in a product group, that they are only in favour of options which would benefit their products, while they are against other options, even if these options have no impact on their product range. The option to add a subcategory for all traditional OFB products (i.e. cider and perry etc.) to the current OFB category would result in legal certainty at EU level and a consistent treatment of borderline products across MS since it would make the current national level non-harmonised distinctions unnecessary.

Section 2: Reduced rates

Reduced rates for small producers

While private individuals and the ‘other’ respondent category tend to be strongly in favour of reduced rates for small producers of all alcoholic beverages, industry stakeholders present more varied positions. Those involved in the production of beer, for instance, strongly agree with reduced rates for small breweries. To the contrary, those involved in the production of wine and intermediate products strongly disagree with reduced rates for small wineries and small producers of intermediate products.

According to stakeholders involved in the beer sector, ensuring that reduced rates are applied also on products from other MS (or third countries) is only a marginal issue. To the contrary, approximately 50% of those involved in the production of spirits consider it a major issue. In addition, determining the independence of a company appears to be a moderate, if not major issue for both beer and spirits producers. All respondent groups expressed agreement (or at least a neutral stance) with the various options aiming at clarifying the rules surrounding small producers (cross-border recognition, certification of independence) and with the ‘no-change’ option.

Industry stakeholders involved in the production of wine, OFB and intermediate products are of the opinion that the introduction of optional reduced rates for small producers in their areas of activity would not reach the goal of increasing small producers’ competitiveness. To the contrary, they fear that this would be the first step for the introduction or the increase of taxes on big producers. The huge majority of stakeholders active in the beer sector maintain that reduced rates are likely, if not very likely, to enhance the competitiveness of small producers, but they agree with the rest of the industry that it may be a tool for MS to introduce/increase taxes for big producers.

Question #33 - Please express your opinion on the following possible approaches to extend the application of reduced rates to small producers of alcoholic beverages that are not currently covered and/or to clarify the implementation rules

100% 1 1 3 5 90% 3 8 8 1 4 3 4 2 9 1 2 80% 10 9 14 3 7 3 6 5 6 5 14 70% 12 7 2 7 60% 2 3 6 23 2 50% 6 3 3 1 14 9 2 40% 14 2 13 8 5 2 1 20 10 4 30% 33 6 9 7 10 20% 5 7 4 2 15 2 18 2 10% 4 1 2 3 2 1 1 0% 1 1 2 1 1 W Ind Priv Oth OFB Ind Priv Oth Int Ind Priv Oth B W OFB Int Eth Priv Oth Introduce in the Directive the Introduce in the Directive the Introduce in the Directive the Clarify the rules on cross-border possibility of reduced excise possibility of applying reduced possibility of reduced excise recognition of small producers duty rates for wine excise duty rates for duty rates for intermediate fermented beverages other products (e.g. sherry, port) than wine and beer (e.g. cider, perry, mead)

100% 1 2 2 1 90% 3 6 1 6 2 2 4 4 7 12 3 8 80% 7 14 6 6 3 70% 2 7 3 8 10 6 7 4 6 14 3 60% 4 50% 2 3 5 8 40% 21 3 8 1 6 3 9 10 7 30% 3 7 9 18 6 7 20% 2 6 3 3 3 2 4 4 10% 1 3 7 5 3 1 2 2 2 0% 1 1 B W OFB Int Eth Priv Oth Eth Ind Priv Oth B W OFB Int Eth Priv Oth Clarify the rules to establish when a producer can be To raise the threshold below No change of the Directive, but the issuance of treated as an ‘independent’ entity which distilleries can be guidelines on the implementation of the rules considered as ‘small’ on reduced rate for small producers

Strongly disagree Disagree Neutral Agree Strongly agree

Legend: B: industry stakeholders with an interest in the beer sector; W: industry stakeholders with an interest in the wine sector; OFB: industry stakeholders with an interest in the other fermented beverages sector; Int: industry stakeholders with an interest in the intermediate products sector; Eth: industry stakeholders with an interest in the ethyl alcohol sector; Ind: rest of the industry (not included in the previous category); Priv: private individuals; Oth: Other (public health NGOs, public authorities, etc.).

Note: industry stakeholders with an interest in the production or end-use of industrial alcohol have been included in the “Ind” category, whenever present. If not present, they have been included in the residual “Oth” category.

In line with the view of the majority of the stakeholders, the preferred options are to clarify the rules when a producer can be treated as 'economically and legally independent' and to introduce an optional reduced rate for small cider producers to improve their competitiveness. As presented in the body of the IA report, extending the optional reduced rate to other producers was, besides undesired, also likely to have limited if any impacts.

Reduced rates for low-strength beverages

As far as wine, intermediate products and ethyl alcohol are concerned, the majority of the industry is against reduced rates for low-strength products. More favourable positions are shown for the other products, especially beer: over 80% of stakeholders in the beer industry are strongly in favour of reduced rates for low-strength beer. Private individuals and other respondents are generally in favour of reduced rates.

The policy option of raising the threshold of low-strength alcoholic beverages – be it moderately or significantly – was faced with strong disagreement by the majority of all respondent groups. The only exception was represented by the beer sector, where almost 80% of brewers would welcome a new threshold for beer set at 3.5% ABV. Opinions on the ‘no-change’ policy option were more varied, with substantial shares of neutral respondents.

Question # 39 - Please express your opinion on the following possible approaches to encourage the use of optional reduced rates for lower strength alcoholic beverages.

100% 1 1 4 2 1 4 1 6 2 5 1 2 90% 4 5 4 1 4 1 4 2 4 3 2 5 2 80% 4 2 15 6 10 1 3 6 10 1 9 70% 3 2 4 1 7 1 4 6 60% 21 4 27 2 4 7 1 50% 4 2 1 2 5 40% 13 2 6 1 2 12 5 12 6 3 23 14 30% 4 32 4 12 23 20% 3 3 15 8 2 2 8 2 13 6 6 7 10% 4 0% 1 B Ind Priv Oth W Ind Priv Oth W Ind Priv Oth OFB Ind Priv Oth OFB Ind Priv Oth To raise the threshold of To moderately raise the To significantly raise the To moderately raise the To significantly raise low-strength beer from threshold of low-strength threshold of low-strength threshold of low-strength the threshold of low- 2.8% to 3.5% ABV wine (currently 8.5% ABV) wine (currently 8.5% ABV) OFB, like cider and perry strength OFB, like (currently 8.5% ABV) cider and perry (currently 8.5% ABV)

100% 1 3 3 1 3 1 1 2 4 2 2 3 3 90% 1 3 4 2 1 4 1 1 4 1 4 2 6 3 1 2 1 5 80% 11 1 9 3 6 4 1 3 6 1 7 2 4 6 5 70% 1 4 1 7 2 3 60% 4 20 4 13 18 5 2 2 50% 2 3 1 4 22 12 2 2 40% 12 4 5 6 1 12 4 11 11 4 1 5 30% 30 9 24 1 20% 2 20 1 8 7 2 7 5 13 5 2 7 2 4 10% 6 4 2 0% Int Ind Priv Oth Int Ind Priv Oth Eth Ind Priv Oth Eth Ind Priv Oth B W OFB Int Eth Priv Oth To moderately raise the To significantly raise the To moderately raise the To significantly raise the No change of the Directive, but threshold of low- threshold of low- threshold of low- threshold of low- the issuance of guidelines on the strength intermediate strength intermediate strength ethyl alcohol, strength ethyl alcohol, implementation of reduced rate products, like sherry andproducts, like sherry and i.e. spirits (currently i.e. spirits (currently for lower-strength alcoholic port (currently 15% port (currently 15% 10% ABV) 10% ABV) beverages, including a clarification ABV) ABV) of the policy objectives

Strongly disagree Disagree Neutral Agree Strongly agree

Note: industry stakeholders with an interest in the production or end-use of industrial alcohol have been included in the “Ind” category whenever present. If not present, they have been included in the residual “Oth” category.

As far as the likely results of reduced rates for low-strength products are concerned, the huge majority of stakeholders involved in the production of beer (and, to a lesser degree, those in the production of OFB) agree that there would be greater incentives for product innovation, with more choice for consumers. Moreover, both small and big producers would benefit, despite a reduction of alcohol consumption per capita. The rest of the industry, however, tend to be far more sceptical on the possible benefits of reduced rates for low-strength products. Unsurprisingly, all respondents – including private individuals and the ‘other’ category, agreed that more affordable low-strength products would not increase the overall consumption of alcohol per capita.

The preferred option is to increase the threshold for low-strength beer from 2.8 % to 3.5 % ABV. This is in line with the preference of the beer industry, while the rest of the industry disagrees with this approach. Similar to the situation as described under section 1, it is clear when the respondents are distinguished by interest in a product group, that they are only in favour of options which would benefit their products, while they are against other options, even if these options have no impact on their product range.

Section 3: Excise duty exemptions for denatured alcohol

Completely Denatured alcohol

The huge majority of all respondents – including those involved in the industrial alcohol sector – stated that in the past 10 years there were no or very few instances of tax frauds involving alcohol sold as potable, even though designated as industrial alcohol.

Over 80% of the industrial alcohol industry, and to a lesser degree also the rest of the industry, agreed that the current legal framework ensured fair competition and flexibility for users of denatured alcohol, facilitating intra-EU trade and helping to fight fraud. Private individuals expressed a somewhat more sceptical opinion. The industrial alcohol industry maintains that the current system is effective and appropriate. The rest of the industry shared the same positive view, with some respondents suggesting that the EU should provide some guidance for the interpretation of the rules. The industrial alcohol industry strongly believes that the distinction between ‘completely denatured’ and ‘denatured’ alcohol is useful and well-defined. At the same time, however, it is of the opinion that the implications of the two categories may not be entirely clear and it is not against the possibility of changing the rules for ‘denatured’ alcohol.

The preferred option to codify the existing regulation is supported by the majority of stakeholders. The current practice is by stakeholders reported as working well and codifying this practice is expected to reduce any remaining trade barriers and distortions and consolidates MS desires for a harmonised solution for CDA into a legal text.

Partially denatured alcohol

In general, a limited number of responses were received on the questions related to PDA. The issues that received greater attention from the industry were: (i) the additional costs and administrative burdens to ensure that alcohol denatured using a formulation accepted in one MS is also recognised in another Member State, and (ii) possible different interpretations on the

meaning of “used for the manufacture of”. Both issues however were described as quite rare, having happened only once or twice.

Industrial alcohol stakeholders strongly agreed with the fact that MS only authorise robust methods for partial denaturation and that they also effectively supervise the production, use and movement of partly denatured alcohol.

The industry of denatured alcohol expressed a strong disagreement with all the policy options that may limit the allowed formulations for the denaturation of alcohol. Moreover, they are against a strict interpretation of Art. 27.1(b). To the contrary, they are strongly in favour of a full mutual recognition, supported by exchanges between public authorities of different MS to better understand each other’s approaches. All the other respondents – including the rest of the industry, private individual and the miscellaneous ‘other’ category – expressed more mixed views on the proposed policy options. As in the previous cases, however, the number of responses from each respondent category is below 10 and cannot be therefore considered truly representative.

Question #53 - Please express your opinion on the following possible approaches to address the problems with the exemption of denatured alcohol.

100% 3 1 3 90% 1 3 1 1 1 80% 3 4 3 1 70% 3 5 5 3

60% 13 1 3 22 50% 3 1 1 1

40% 1 21 3 1 19 1 2 30% 1 1 1 20% 2 4 8 1 2 1 1 1 10% 2 1 1 2 2 1 2 0% IA Ind Priv Oth IA Ind Priv Oth IA Ind Priv Oth IA Ind Priv Oth Agree on the “Euro” denaturant (in Insist on full mutual recognition of Accept limited mutual recognition of Develop a common list of whichever concentration) as the only completely denatured alcohol, to completely denatured alcohol, meaning formulations that are authorised formulation under article ensure that all Member States recognise that each Member State is only obliged authorised for denatured 27.1 (a), and the eventual elimination of all procedures notified by all Member to recognise alcohol denatured using its alcohol under 27.1 (b) across all all national procedures States, irrespectively of where the own national method(s) (which may Member States, distinguishing alcohol was produced / denatured include the "Euro" denaturant) between different end-uses / sectors

100% 1 1 1 2 90% 1 1 1 2 2 1 1 1 1 3 1 80% 3 4 70% 4 3 3 2 5 60% 4 2 1 3 22 50% 2 1 1 4 24 23 22 2 40% 2 1 2 2 30% 1 1 1 2 20% 1 1 2 2 1 1 2 1 10% 2 2 1 1 1 1 1 2 1 1 0% IA Ind Priv Oth IA Ind Priv Oth IA Ind Priv Oth IA Ind Priv Oth IA Ind Priv Oth Clarify that the terms “used for Clarify that the terms “any Fund capacity / confidence Define conditions for the Depart from the current the manufacture of” in article product not for human building measures (e.g. training, Member States to apply Articles distinction between 27.1 (b) only refers to the direct consumption” in article 27.1 (b) exchanges, visits) between 30 and 31 of Directive “completely denatured” use, and does not include requires that the product in national tax and/or customs 2008/118/EC for the movement and (partly) “denatured” indirect uses such as cleaning, question is a recognisable authorities to enhance their of products of low fiscal risk to alcohol; and instead disinfection or other adjacent finished product, and held out understanding of each others’ certain types of denatured distinguish between activities for sale as such approaches to supervising the alcohol different end uses (e.g.: production and movement of (1) sale to the public; (2) denatured alcohol industrial applications; (3) use in biofuels)

Strongly disagree Disagree Neutral Agree Strongly agree

The preferred option to clarify that ''any product not for human consumption'' makes reference to a 'recognisable finished product' / 'finished product' is not supported by the majority of the industry involved in the production or end-use of alcohol for industrial uses. It should be noted, however, that the reason this area is under review, is to limit fraud and the way Member States authorities' can combat and detect fraud in this area. This preferred option is the only way to

provide MS with flexibility for the various product groups using PDA. Moreover, it would ensure equal treatment of goods containing PDA across the EU and reduce the risk of costs associated with disputes between businesses and national authorities.

The different national approaches towards PDA which are currently in place hamper the smaller businesses in cross border trade or even to enter the market in general, because of the investment required. This situation is in some way 'protecting' the current operators on the market, which is undesired from an internal market and competition perspective. Unsurprisingly, the current operators, managing to deal with all difficulties of understanding and complying with the different frameworks and costs, desire the status quo. The functioning of the internal market and the fight against fraud have been given more weight than to ease the way of operating of the current industry, hence the choice for an option which does not correspond to the view of the industry involved in the production or end-use of alcohol for industrial uses.

Section 4: Calculation of excise duties on sweetened or flavoured beer using the Plato method

The majority of respondents, including those involved in the production of beer, believe that the term “finished product” in Art.3.1 should be interpreted as the end product, after the addition of sweeteners and flavourings. It is worth noting, however, that big beer producers and many associations of breweries decided not to answer the question, clarifying their position in the additional comments. They explained that, while the Directive makes reference to the Plato method, it does not clearly define how to measure the Plato degree. For this reason, they suggest applying the common and ‘everyday’ understanding of the brewing industry relying on the Balling equations, which does not consider sweeteners and flavourings added after fermentation.

While private individuals would be in favour of a revision of Article 3.1 to clarify the meaning of finished product, the beer industry tends to disagree. The rest of the industry, whose response rate was rather low, expressed varied positions on the issue. To the contrary, the majority of the industry – both involved and not involved in the production of beer – agreed or strongly agreed with the ‘no-change’ policy option, with guidance on the ‘correct’ approach to measure the Plato degree.

Question #56 - Please express your opinion on the following possible approaches to the issue of excise duty applicable to sweetened or flavoured beer measured by degree Plato.

100% 90% 2 9 3 80% 2 70% 20 7 3 2 9 1 60% 2 50% 2 2 3 1 40% 3 30% 16 2 20% 5 4 6 2 5 10% 2 1 0% 2 1 2 B Ind Priv Oth B Ind Priv Oth Clarify article 3(1) of the Directive, with respect to the No change to the Directive, but guidance on the interpretation of “finished products” definition ‘correct’ approach to measure degree Plato of sweetened/flavoured beer

Strongly disagree Disagree Neutral Agree Strongly agree

Legend: B: industry stakeholders with an interest in the beer sector; Ind: rest of the industry (not included in the previous category); Priv: private individuals; Oth: Other (public health NGOs, public authorities, etc.).

The preferred option to clarify the calculation of excise duty on sweetened or flavoured beer using the Plato method is to amend Article 3 of the Directive. This is the only option to provide legal clarity and ensure an equal approach of the calculation of excise duty using the Plato method across MS. To avoid disrupting the functioning of the internal market as much as possible, the preferred option is to codify the current practice in the majority of MS using the Plato method.

5. REFIT Platform opinion

For completeness, it is worth summarising the main points of a recently issued REFIT Platform Opinion that addressed some of the present issues38. The Opinion concerned the common interpretation of EU laws on Wine and Spirits.

The Stakeholder group recommended the Commission to reduce the room for interpretation in the Directives on wine and spirits by issuing a common threshold on fermented alcohol. The Stakeholder group further suggested that the Commission would ensure harmonization in the taxation on wine and spirits in the EU and national measures to eliminate room for national interpretation. In particular, it emerged that while various Member States support more accurate definitions and greater clarity in legislation in order to reduce legal uncertainty, views are divided on whether this should be achieved by establishing common thresholds on the amount of fermented alcohol used in mixtures.

More generally, the debate also showed the persistence of divided views on the structure of excise duty, with some MS in favour of taxation per alcohol content, while for other MS the current rules should not be touched.

38 REFIT Platform Opinion on the submission by the Scottish Council for Development and Industry and a Member of the REFIT Platform Stakeholder group on the common interpretation of EU laws on Wine and Spirits Date of Adoption: 07/06/2017. https://ec.europa.eu/info/sites/info/files/xviii12abinterpretationeulawswine.pdf

ANNEX 3. WHO IS AFFECTED BY THE INITIATIVE AND HOW

1. Practical implications of the initiative

National authorities are affected by this initiative as it will increase the legal certainty and thereby reduce their administrative burden.

Economic operators who manufacture or use alcohol are affected by this inititave. This initiative reduces the legal uncertainties that currently exist and reduces the competitive distortions between economic operators who are involved in cross-border trade.

The costs and benefits of each option are analysed in detail in section 6 for each distinctive problem area.

2. Summary of costs and benefits

The tables below summarise the costs and benefits (in million euros) for all preferred options together.

I. Overview of Benefits (total for all provisions) – Preferred Options Amount Description Comments per year Direct benefits Member States Introducing a differentiation in the OFB category will No reduce the instances of BTI shopping, laboratory change39 testing and administrative burden to deal with these

products. Administrative costs savings

Clarification of independent brewer will ease the cost

of implementing the reduced rates scheme for Member

States.

Less scope for intentional misclassification of PDA so as to avoid controls Fiscal fraud and associated revenue, ++ health risk Addressing the classification of certain alcoholic beverages will reduce the foregone tax revenues of products that exploit the current uncertainties.

Clarification will ensure a harmonised approach across the EU in terms of indirect uses of denatured alcohol, measurement of the Plato degree of sweetened /flavoured beers, classification of certain alcoholic Legal Certainty +++ beverages and the operation of the reduced rates for small producers.

This clarification will reduce the divergent interpretations, disputes and associated costs for

39 No change, as in the worst case scenario benefits could be neutralised by the risk of new borderline products

Member States.

Economic Operators

Cost savings for users of PDA in MS that do not currently exempt indirect uses. Clarification of PDA terms will lower the risk of delays / costs associated with disputes with authorities.

Operating costs + Clarification on classifying certain alcoholic beverages will reduce the Clarification / extension of the scheme for small producers will result in costs savings for small brewers / cider makers Indirect benefits Economic Operators

Clarification will reduce the divergent interpretations and associated costs for economic operators. Greater transparency and legal certainty may result in fairer competition and improve the ease of doing Functioning of the single market and business cross-border. competition The competitiveness of small producers and low strength alcohol would be greatly enhanced, as diseconomies of scale and market access barriers could be counterbalanced.

II. Overview of costs (million euros) – Preferred options Businesses Administrations

One-off Recurrent One-off Recurrent Direct costs 0.00 Minimal 0.00 Minimal Denatured alcohol Indirect costs 0.00 0.00 0.00 0.00 9.2- Classification of certain 35 - Direct costs 15.640 0.00 +42 alcoholic beverages 25043 1.041

40 Administrative costs of a new category 41 Administrative burden from a new EPC 42 Impact of options on competent authorities cannot be quantified. The affected population is limited so in aggregated terms it may be modest. 43 Lost tax revenues due to a new category - the gap ranges from EUR 35 million if borderline cider is kept out of the reclassification process.

Indirect costs 0.00 Modest44 0.00 0.00 Direct costs 0.00 0.4545 Modest Modest Reduced rates – small brewers Indirect costs 0.00 0.00 0.00 0.00 Minimal EUR Minimal 15 Reduced rates – small cider Direct costs makers 0.32 / hl Indirect costs 0.00 0.00 0.00 0.00 0.00 0.00 0.00 <1% of Reduced rates – low strength Direct costs total tax beer revenue Indirect costs 0.00 0.00 0.00 0.00 Direct costs Minimal Minimal Plato Indirect costs 0.00 0.00 0.00 0.00 Direct costs TOTAL Indirect costs

3. EVIDENCE AND STAKEHOLDERS' VIEWS ON THE ADMINISTRATIVE COSTS AND BURDENS

Revision of Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages is part of the Commission's REFIT programme. Evidence collected in the evaluation and feedback gathered from the day-to-day application of the Directive's provisions led to the conclusion that there was a perceivable – albeit hardly quantifiable – lack of legal certainty over the treatment of specific products, leading in turn to potential additional costs to economic operators.

The resulting room for diverging interpretation involuntarily allowed since 1992 economic operators, as well as national tax administrations, to establish their own modus operandi. Although it could be concluded after the Ramboll Evaluation and the Study that the Directive functions generally well overall, the diverging interpretations result in problems where these views differed. Most of the problems identified – as described in the problem definition section – were specific to certain markets or certain products. In terms of REFIT objectives, these focus particularly on those areas where economic operators see excessive costs and burdens. In the case of Directive 92/83/EEC the discussion is shifted from not so much the excessive costs and burdens to unnecessary costs and burdens, which could be avoided if the Directive functioned better.

For example, the majority of problems relate to the legal uncertainty that the economic operators experience with production, use and/or movement of some alcoholic products. These are the costs that stem not from the requirements of the Directive as such (most of the requirements related to the holding and movement of excisable products stem from Directive 2008/118/EEC) but from its imperfect application or, in certain cases, the fact that the legislation has not kept abreast of new developments.

44 Some aromatised wine products classified as CN 2206 may fall within the reclassification, (the estimated decline in sales when compared to the overall alcoholic beverage markets is less than 0.4% in the worst scenario) 45 Administrative burden linked to the uniform certificate for small brewers.

Summary of REFIT costs reduction

with authorities of the were not possible to estimate

receiving MS

PROBLEM 1 cross-border equal treatment of goods elimination thereof- are case-

Overall burden not expected to change significantly (€ 2.0 – Legal certainty at EU level 2.5 million)

Classification PROBLEM 2 the benefits in 10 years or more neutralise benefits of any new clear definition IT updates: approx. €800 per company or aggregated burden of to € 6.9 million.

Recognised small brewers: total burdens for 675 operators in the sample MS: approx. In terms of recognising the €13 000 or 2% of the

Plato degree PROBLEM 3 across the EU elimination

ANNEX 4. ANALYTICAL MODELS USED IN PREPARING THE IMPACT ASSESSMENT

The bulk of the data collection work was centred on a vast in-depth consultation of stakeholders, covering a total of 12 Member States, as well as EU-level institutions and organisations. Overall, 160 interviews were conducted with different types of stakeholders, namely: public authorities and administrations (tax and customs authorities, public health authorities, agriculture authorities and others); economic operators of different size and active in different segments of the market and the value-chain; non-government public health organisations; and various other alcohol market experts. The interview programme was complemented by an Open Public Consultation that received a total of 166 responses.

The Study results are also based on the result of the econometric analysis of an extensive database, with EU market data on volume, value and price of some 1,374 brand lines in the 1990-2016 (including forecasts until 2021) period, as well as on a comprehensive desk research including: EU and MS-level policy documents, scientific literature, various institutional databases, industry and stakeholder reports and papers, web-sources and other grey literature, both published and unpublished.

The main focus of the analytical work was to compare the ‘no change’ scenario, developed on the basis of an in-depth baseline assessment, with several ‘policy change’ scenarios, using both quantitative (cost/benefit) and qualitative (multi-criteria) methods. The impacts considered for the comparison of scenarios belong to four main categories: (i) tax revenues and burden; (ii) regulatory costs and cost savings (including substantive compliance costs, administrative costs and enforcement costs); (iii) market effects (including Single Market functioning, distortion of competition, and SME competitiveness effects); and (iv) indirect social effects (illegal activities and fraud, and alcohol control policy objectives).

The proposed policy options for the revision of the Directive may determine a variety of different economic and social impacts for various different stakeholder groups, primarily MS competent authorities and economic operators, secondarily consumers and public health stakeholders. The different typologies of impacts assessed in this Study can be gathered in five main categories, as follows:

Direct charges. Direct charges include taxes and fees paid by economic operators or consumers. In line with the nature and scope of Directive 92/83/EEC, the focus of this Study is excise duty on alcohol and alcoholic beverages, and the related excise duty revenues of Member States. This dimension has been examined across all thematic areas considered. Unless differently stated, all references to ‘tax rates’, ‘tax structures’, ‘tax revenues’ etc. in this Report relate to excise duties. However, in some cases, the analysis has also encompassed VAT, and in particular the share of VAT that is imposed on the excise duty since this causes a multiplier effects on the variation of excise duty levels.

Importantly, tax revenues have distributional impacts: what is a benefit for tax authorities may be a cost for consumers and/or manufacturers. In the assessment and comparison of policy scenarios these impacts where primarily examined from the perspective of tax authorities. In this sense an increase of tax revenues is rated positively and vice versa. Impacts on tax revenues can be triggered by variations of: (i) rates applicable to excisable products, which is outside the scope of the Study, with the exception of provisions on reduced rates or methods for the calculation of applicable tax; and (ii) scope of the tax system (exemptions / inclusions) and of individual tax category, with the possible re-classification of certain products in different categories. It is also worth mentioning that these variations may also trigger other impacts, considered below under market or social effects, such as tax-induced substitution between products, per capita consumption effects, demand for illicit products and fraud.

Compliance, administrative burden, costs and cost savings. Compliance costs have been considered with respect to the changes to business practices linked to the administrative requirements concerning denatured alcohol. Administrative burden for economic operators have been assessed in various policy options implying a revision of the Directive. For instance, the creation of a new fiscal category for certain products may generate administrative burden for economic operators, who have to update their licenses and IT systems. Another example is the quantification of the current administrative burden generated by the reduced rate scheme for small brewers, and the costs associated with its revision, or to the extension of the scheme to small wine producers and cider makers.

Enforcement costs and benefits. As regards enforcement costs and benefits, two main types have been considered:

(i) enforcement costs and cost savings stricto sensu, which are those borne by public authorities to apply the revised Directive provisions; and (ii) judicial costs and cost savings, which are costs borne by public authorities and economic operators related to the need to interpret unclear legal provisions and, in case of judicial disputes, uphold them in court, as well as benefits (cost savings) where interpretations and judicial disputes are no longer needed after a clarification or legal revision.

Market effects. Market effects concern distortions of the quantity exchanged and of the equilibrium price of the various products. Taxation, by definition, distorts any market from the equilibrium that it would reach based on the free adjustment of demand and supply. For this reason, the Study did not attempt to assess market distortions per se, but those that might go beyond the intended objectives of the regulator, in terms of Single Market functioning. Four categories of possible market effects and distortions have been considered:

1) Tax-induced substitution across products, i.e. when the demand for a certain product is favoured (hampered) by the higher (lower) taxation imposed on one or more substitute products. 2) Cross-border distortions and illicit markets. This may be the case when consumers decide to purchase a certain product (e.g. alcoholic beverages, denatured alcohol) in another MS, or stop importing the same product from another MS, because of the different tax or regulatory treatment. Effects on illicit (or informal) markets also include impacts from and on the quantity of ‘unrecorded alcohol’, i.e. alcohol which is not taxed and is outside governmental control. This includes cross-border shopping (both legal, and smuggling / ‘bootlegging’) and, most importantly for the policy issues considered, surrogate products obtained from previously denatured alcohol and home production. 3) Single Market functioning, and possible distortions induced by diverging legal treatments or uneven application of Directive provisions or other administrative obstacles hampering the circulation of products or affecting fair competition. 4) SME competitiveness, since certain impact may have a differential effects on small producers versus large manufacturers. This is specifically the case for the analysis of the ‘reduced rates for small producers’ issue, where both the baseline assessment and the impact analysis consider whether and to what extent the norms at stake change the competitive position of SME.

Indirect social effects. This category includes impacts that poorly lend themselves to a quantification in monetary terms, but are nonetheless important since they concern the underlying values and principles of policy action that are linked to social well-being in the broad sense. Two areas of social impact that have been considered related to the policy options at stake - although

indirectly - are namely: (i) public health (through alcohol control policy and measures); and (ii) tax fraud.

The final step of the analysis of impacts consisted of the comparison of the policy options. The issues at stake in this Study require policy revisions that are relatively independent from one another. Therefore, the comparison of options have been performed for each thematic area separately, rather than in a cumulative way. Given the different nature of the impacts considered, the final comparisons required combining different approaches, and specifically, a partial cost- benefits analysis (CBA) approach for quantifiable (monetary) impacts, such as market effects, tax revenues and – where feasible – regulatory costs, and a multi-criteria analysis (MCA) for non- quantifiable or mixed ones.

ANNEX 5. COUNCIL CONCLUSIONS (6 DECEMBER 2016)

The Council conclusions on the Commission Report to the Council on the evaluation of Directive 92/83/EEC, adopted by the Council at its 3506th meeting held on 6 December 2016.

COUNCIL CONCLUSIONS ON THE COMMISSION REPORT TO THE COUNCIL ON THE EVALUATION OF COUNCIL DIRECTIVE 92/83/EEC ON THE STRUCTURES OF EXCISE DUTIES ON ALCOHOL AND ALCOHOLIC BEVERAGES

The Council (ECOFIN):

1. WELCOMES the Commission Report to the Council on the evaluation of Council Directive 92/83/EEC on the structures of excise duties on alcohol and alcoholic beverages and TAKES NOTE of the findings and recommendations set out in that Report.

2. AGREES with the assessment that, in general, Directive 92/83/EEC functions effectively and makes it possible to avoid tax-related trade barriers or competitive disruptions between economic operators in the same sector of activity.

3. TAKES NOTE that the Commission Report concentrates exclusively on the structures of excise duties on alcohol and alcoholic beverages, and in no way either covers or combines the findings with the requirements laid down in Directive 92/84/EEC on the approximation of the rates of excise duty on alcohol and alcoholic beverages.

4. CONFIRMS, that it is necessary to prevent ambiguities leading to distortions of competition between economic operators and to apply harmonised conditions and rules for taxing alcohol and alcoholic beverages. Moreover, it is essential to provide equal conditions for economic operators in the functional internal market, eliminate disruptions to fair competition and prevent tax evasion and avoidance.

5. NOTES, however, that the Directive could be amended as appropriate in order to eliminate certain ambiguities that sometimes cause particular types of alcohol and alcoholic beverages to be treated differently. This would also improve collection of excise duties and reduce administrative costs for both economic operators and tax administrations in Member States.

6. RECOGNISES the need to clarify and to harmonise further the classification rules for products manufactured as mixtures of different categories of alcoholic beverages or as mixtures of alcoholic beverages with non-alcoholic beverages in order to unify the treatment for excise purposes of the same products across the Member States, and so ensure legal certainty and clarity for economic operators.

7. STRESSES the need to ensure uniform treatment of alcoholic beverages, which are the mixture of fermented beverages and alcohol, and in this context, for the purposes of legal certainty, to clarify the notion of “entirely of fermented origin” in Directive 92/83/EEC.

8. AGREES with the assessment that clear rules are in place for applying reduced rates for small producers of beer and ethyl alcohol and INVITES the Commission to investigate the impact of extending those rules to small producers of still and sparkling wines, other fermented beverages and intermediate products.

9. INVITES the Commission to investigate further the potential impact of allowing Member States to exempt from excise duties the production of ethyl alcohol and intermediate products for own consumption and to present a report to the Council and RECALLS the particular importance of striking the right balance between revenue, the costs of tax administration, other aspects relating to consumption and the impact on cross-border trade.

10. TAKES NOTE of the recent adoption of Commission Implementing Regulation 2016/1867/EU stipulating one common "euro" procedure for completely denaturing alcohol and in this context, RECOGNISES that Article 27 of Directive 92/83/EEC, more generally, needs to be updated, in order to define the transparent and clear conditions for applying the exemptions for any type of denatured alcohol, without prejudice to the Member States' competences.

11. RECALLS the need to achieve the right balance between preventing tax evasion and avoidance while ensuring flexibility in the use of different denaturing procedures laid down by the Member States in accordance with the point (b) of Article 27 (1) of Directive 92/83/EEC, and ENCOURAGES the Commission to develop, in collaboration with all Member States, a clear definition of final products, which should eliminate the consequences of different treatment of products from denatured alcohol within internal market.

12. TAKES NOTE that in order to ensure further harmonization of the exemptions provided for in points (a) and (b) of Article 27 (1) of Directive 92/83/EEC, it might be necessary to amend the rules on holding and transporting denatured alcohol to reflect the provisions of Council Directive 2008/118/EC.

13. RECOGNISES that some CN codes referred to in Directive 92/83/EEC need to be updated, as this Directive was adopted more than 20 years ago.

14. TAKES NOTE that, in the interest of clarity and given the potential revision of Directive 92/83/EEC, rules that were designed for specific Member States and are no longer used could be removed.

15. REQUESTS that the Commission, taking into account these Council Conclusions and the objectives set out in Directive 92/83/EEC, undertakes all relevant studies and, after carrying out the relevant technical analysis, public consultations and an impact assessment, submits to the Council an appropriate legislative proposal in 2017 or, in case it chooses not to submit a proposal, informs the Council of the reasons.

100

ANNEX 6. DRIVERS OF THE DYSFUNCTIONS IN THE APPLICATION OF EXEMPTIONS FOR DENATURED ALCOHOL

Driver 1: an incomplete / inconsistent mutual recognition of CDA

According to the Directive, MS are allowed to use their own methods for complete denaturing of alcohol as long as they notify them to the Commission. These formulations are to be further mutually recognised by all MS. Until recently, the majority of the MS had at least one national denaturing formulation notified, in addition to the EU common denaturing method known as the Eurodenaturant. The extent to which MS indeed respected the principle of mutual recognition varied considerably.

The problem was not so acute in the context of the domestic market but it became potentially distortive when CDA was moved over borders, from one MS to another. In the Study, some MS specifically stated that they would only recognise a method authorised by the MS of origin of the denatured alcohol (e.g. denatured alcohol from Slovakia using Germany’s recognised method of complete denaturation would not be recognised as a CDA in Romania); other MS held the view that producers from other MS selling CDA to that country would be able to choose from all the different formulations notified; finally a few MS would only accept their own denaturing method or the Eurodenaturant.

Paradoxically, producers of denatured alcohol in third countries enjoyed in practice greater freedom and opportunities, with most of the MS recognising any denaturing formulation notified in accordance with the Directive.

For CDA, the adoption of the new Commission Implementing Regulation 2017/223646 has clarified the position for most MS by the implementation of a common harmonised CDA formulation. However, in order to truly complete the harmonisation, this needs to be reflected in the Directive because as it is currently written, there is a significant risk that it could attract the re-introduction of weaker CDA formulations via the notification procedure, and undo all the work to date.

Driver 2: divergent national approaches to partially denaturing alcohol (PDA)

In terms of PDA, the EU rules are largely non-harmonised and there are thousands of PDA formulations in use across the EU. The European alcohol denaturant database47, which holds information of formulations, is currently out of date, not sufficiently maintained by the MS administrators and there is no external access to industry to allow them to check the validity of a particular formulation.

There are fundamental differences in the control and administration regimes applied in MS, creating unfair competition and burdens on business. The Directive only stipulates that PDA is exempted from excise duty while leaving the modalities regarding partial denaturation to the MS. Moreover, unlike in the case of the CDA, the duty exemption is conditional on the basis that it has been used in the manufacture of a product which is unfit for human consumption. Some PDA is moved cross borders under the duty suspension rules using Excise Movement &

46 Commission Implementing Regulation (EU) 2017/2236 of 5 December 2017 amending Regulation (EC) No 3199/93 on the mutual recognition of procedures for the complete denaturing of alcohol for the purposes of exemption from excise duty, OJ L 320, 6.12.2017, p. 6–9. 47 This database is managed by the Commission (Joint Research Centre).

101

Control System (EMCS), which is a computerised system for monitoring the movement of excise goods under duty suspension in the EU, which creates a financial and administrative burden on the economic operator. Whereas in other MS economic operators are allowed to release the “finished” product (often in bulk, e.g. screenwash), which is in free circulation (i.e. EMCS is not used) and is not considered an excise product anymore. In this scenario, the MS do not necessarily need to recognise each other's denaturing formulations.

MS’ regulatory and administrative frameworks vary significantly regarding the procedures governing supervision of the production, movement and use of PDA, formulations they have authorised (sometimes just a few, sometimes hundreds) and the process for obtaining authorisations (in some MS this is limited to the formulations on the official published list, whereas others can authorise formulations ‘ad hoc’ for individual economic operators).

The PDA made in a given MS in accordance with its national requirements can be moved to another MS using the EMCS. However, it will not be recognised as legally denatured (and thus not exempted from excise duty) when used for the manufacture of products, unless it also complies with the formulation and authorisation requirements of the receiving MS (which may be different). This frequent lack of transparency of formulations authorised in each MS was described by producers as difficult and time consuming. German producers advised that is not viable to supply customers in some MS (such as Czech Republic, France) as the supervisory regimes in these MS are particularly strict towards foreign producers.

While the use of EMCS is obligatory for all cross-border movements, some (but not all) MS allow simplified procedures for movements of PDA within their own territory. Furthermore each MS has their own system of registrations, licenses and authorisations for alcohol producers and economic operators using alcohol in the production process. Spain requires the presence of a tax official during the denaturation process. France allow for "in situ" denaturation in certain sectors (cosmetics), meaning users can buy pure alcohol and denature it on their own premises, sometimes as part of the production process. These variances cause problems for industry as it is described in detail under consequences (see section 2.3.5).

The greater problem is the difference in application among MS as to the uses of PDA, and how the Directive is ambiguous in its attempt to define denatured alcohol which is used in the manufacture of a product not for human consumption. Additional to this, there are thousands of PDA formulations in use across the EU and many of them do not contain chemical analytical markers, rendering them impossible to detect in illicit potable spirit. Consequently this leads to very strong enforcement and compliance regimes in some MS – which puts financial and administrative burdens on some economic operators in these MS. This is best illustrated in the cosmetics and perfumes sector. Manufacturers who operate in several MS find themselves making the same products using PDA but where the holding, movement and even the processing of the alcohol is different. It is clear from the Ramboll Evaluation and from the outputs of the Fiscalis Project Group that in some circumstances, MS have found some practical workarounds to the problems for PDA, but this has not overcome the inconsistency problem of application across the EU nor has it addressed the concerns of industry that there is no level playing field in certain manufacturing sectors.

Overall, the vast majority of national authorities and economic operators alike, consulted in the context of the supporting Studies, appreciated the flexibility offered by the system, leaving it to the MS to apply rules that best meet the needs of their industry (which vary significantly from sector to sector), while reducing the risk of fraud to a level that is deemed acceptable by the MS in question. The single market functions properly for products containing PDA, in so far that, for example, perfume made in France containing PDA in accordance with the French requirements can be sold freely across the EU. At the same time, it is clear that the same is not entirely the case for PDA itself because of the lack of harmonisation.

102

Driver 3: divergent interpretations of certain terms related to PDA

The ambiguous text of Article 27(1)(b) which defines PDA and related terms continues to cause uncertainties and discrepancies. The term ‘used for the manufacture of’ includes indirect uses (such as cleaning manufacturing equipment and production lines). However a minority of MS do not consider that PDA used for these purposes qualifies for the exemption. This results in an unfair treatment across the EU and costs for users in these MS (i.e. denatured alcohol used in cleaning lines which is then destroyed attracts the same excise duty rate as potable alcohol.)

As noted previously, PDA itself has to be moved under the duty suspension regime, whereas finished products containing PDA are released for free circulation. Thus, the question of when alcohol ceases to be classified as PDA, and becomes a finished product, has important practical consequences. There are disagreements as to what constitutes a ‘finished product’ containing PDA that can be exempted from excise duty and released for consumption. Doubts can arise in particular regarding products with very high alcohol content, such as screenwash or other cleaning products. This creates legal uncertainty and increases the risk of fraud as the scope for the misclassification of PDA mixed with very small quantities of other substances is possible. (This is discussed further under driver 4, which deals with fraudulent uses of PDA.)

It should be noted that the Indirect Taxes Expert Group adopted an opinion in 2014 to clarify the term finished product, however as the recommendation is not legally binding, some diverging national practices continue.

Driver 4: potential for fraudulent use of denatured alcohol

Although comprehensive and reliable evidence is not well documented, there are strong indications that, in some MS at least, fraud with denatured alcohol is significant. This risk is associated with the diversion of alcohol intended for industrial uses into the potable alcohol market. Stakeholders predominantly in Eastern European MS (including CZ, PL, LT) reported that fraud involving denatured / industrial alcohol is a significant concern. From a public health perspective, certain denaturants (in particular methanol, which is widely considered the greatest hazard) are toxic, and can lead to illness and even death when consumed.

The role of surrogate alcohol (i.e. purified denatured alcohol) within this is also likely to vary significantly. In many MS (including DE, ES, FR), the interviewed stakeholders unanimously agreed that the consumption of surrogate alcohol is almost unheard of due to a combination of cultural and socioeconomic factors. These factors include the comparatively low levels of excise duty for alcoholic beverages, (including the zero rate on wine) meaning legal alcohol is cheaply available, which reduces the incentives for fraud and makes the purification of denatured alcohol unattractive economically). However a recent seizure worth EUR 460 000 in Ireland indicates that fraud involving denatured industrial alcohol occurs in other MS as well48.

In the responses to the 2015 Ramboll evaluation questionnaire, the highest estimates (from a small minority of MS) were that abuses of the exemptions for denatured alcohol were responsible for 40-80% of the loss of spirits duty from fraud.

The manifestations of the problem are varied. They include purified denatured alcohol (typically from solvents, thinners, barbecue firelighters, screenwash or anti-freeze) from which the smelling and/or tasting agents have been chemically removed, and which is then used for the

48 https://www.revenue.ie/en/corporate/press-office/press-releases/2017/pr-301117-alcohol.aspx

103

manufacture of illicit beverages (usually spirits). There are also reports of cosmetics or personal care products, such as mouthwash or after shave, that can be drunk ‘as is’ (i.e. without removing the denaturants), and are sometimes sold and bought with this purpose in mind. Currently there is a Court of Justice of the European Union (CJEU) case (C-567/17) regarding alcohol which was partially denatured in accordance with the requirements of a MS and subsequently the alcohol was moved, duty exempt, to Lithuania for use in the production of cosmetic products, including mouthwash.

Box 1 – CJEU case C- 567/17: Bene Factum’ UAB (Lithuania)

The Lithuanian company is engaged in the manufacture of and trade in cosmetics and personal care products. The Lithuanian company bought mouthwashes and cosmetic alcohol from a Polish company for commercial purposes. The ethyl alcohol contained in the products was denatured in accordance with the requirements of Poland before the products were transported to Lithuania. Accordingly, the Lithuanian company, relying on the provisions of the Directive, considered the ethyl alcohol contained in the products exempt from excise duty.

The Lithuanian tax authorities carried out a tax inspection, which indicated that the products were supplied to various wholesale and retail undertakings and ultimately the products were sold as intoxicating alcoholic beverages. The tax authorities believe that the Lithuanian company failed to take any real action to prevent these cosmetic products being consumed as alcoholic beverages and therefore believe that the products are liable to excise duty.

Both parties agree that the ethyl alcohol contained in the products was denatured in accordance with the requirements of a MS. However they disagree as to whether or not the products meet the condition that they are not fit for human consumption. The CJEU is called to rule on whether the exemption in article 27 (1)(b) should be interpreted as applying to any products not for human consumption in accordance with their basic (direct) intended use, irrespective of the fact that some individuals may consume cosmetic products as alcoholic beverages for intoxication purposes.

Another way the problem manifests itself involves 'finished products' containing alcohol (whether denatured or not) that are classified and shipped as something else (see driver 3 above). Several MS reported being aware of cases where bulk shipments of alcohol with only minimal quantities of other ingredients (such as detergent and/or colour) were declared as Combined Nomenclature (CN) codes other than 2207 – such as CN codes 3820 (anti-freeze) or 3824 (miscellaneous chemical products) – and therefore moved without any controls. National authorities admitted that they could not be certain of the scale of the problem, since such movements are not registered under EMCS, and detection therefore relies on more or less random checks. Some MS thought there were only a few isolated cases, but other MS (especially CZ, PL) believed it could be significant and provide a route for fraudulent activity.

The risks are obviously proportionate to the cost of the products in question – for example, many stakeholders tend to agree that the high retail cost of certain perfumes justifies the use of ‘weak’ denaturants (such as essential oils), as there is no risk of such products being purchased with the intention of drinking them. Others, however, insist on the importance of including at least a chemical marker in all PDA formulations as a matter of principle.

104

ANNEX 7. CLASSIFICATION OF ALCOHOLIC BEVERAGES

1 Customs classification

The relevant classification for trading purposes of alcohol and alcoholic beverages is the customs classification. Laid down in the Combined Nomenclature (CN)49 - a further development of the Harmonized System (HS) nomenclature of the World Customs Organization (WCO)50 - this classification is used to determine the applicable tariff (‘tarification’) to goods declared to customs in the EU. As discussed further below, the CN classification determines also the excise duty category of products and is therefore at the core of the ‘classification issue’ described in this Section.

CN codes have (on CN level) 8-digits. The first 4-digits are the most important, since they define the product ‘heading’ and are relevant for the determination of the excise duty. In a few cases, however, the tax categorisation of certain products make reference to 6-digit or 8-digit sub-headings (e.g. for sparkling wine and other fermented beverages). As of the latest revision51, the customs classification included an overall 180 8-digit sub-headings clustered into six main headings as in Table 4 below.

Table 4 – Structure of the HS/CN classification of alcohol and alcoholic beverages

CN / HS headings (4 digits) 6-digit52 8-digit

2203 Beer made from malt. none 3 subheadings 2204 Wine of fresh grapes, including fortified wines; grape 5 126 subheadings must other than that of heading 2009. subheadings 2205 Vermouth and other wine of fresh grapes flavoured 2 4 subheadings with plants or aromatic substances. subheadings 2206 Other fermented beverages (for example cider, perry, mead); mixtures of fermented beverages and mixtures of none 7 subheadings fermented beverages and non-alcoholic beverages, not elsewhere specified or included. 2207 Undenatured ethyl alcohol of an alcoholic strength by volume of 80% vol or higher; ethyl alcohol and other spirits, none 2 subheadings denatured, of any strength. 2208 Undenatured ethyl alcohol of an alcoholic strength by 7 volume of less than 80% vol; spirits, liqueurs and other 54 subheadings subheadings spirituous beverages. Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017; author’s analysis of EBTI database (accessed in February 2017).

The CN, and its parent HS, are closed systems designed to comprehend all traded products, so each heading includes one or more residual ‘other’ category(ies) to cover products not explicitly mentioned in the definitions. This entails that new products not strictly matching the definitions

49 http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ:L:2016:294:TOC 50 See http://www.wcoomd.org/en/topics/nomenclature/instrument-and-tools/hs-nomenclature-2017-edition.aspx 51 Commission implementing regulation (EU) 2016/1821 of 6 October 2016 amending Annex I to Council Regulation (EEC) No 2658/87 on the tariff and statistical nomenclature and on the Common Customs Tariff, OJ L 294, 28.10.2016. 52 Until the 6-digit level the CN and the HS codes coincide.

105

provided should in any case fit into one of the existing CN codes. To facilitate coding, the CN (and the HS) is underpinned by non-binding Explanatory Notes (CNEN), which are revised and adjusted periodically.

For legal certainty on the correct ‘tarification’ of beverages, and to prevent the risk that the attribution of a certain CN code is challenged (and fined) by customs or tax authorities when the product is already commercialised, economic operators may apply for a Binding Tariff Information (BTI). These are classification decisions issued by the customs administration of any Member State, which are binding throughout the EU for a period of normally three years (unless the classification code changes or it is affected by EU or international customs tariff measures or by a CJEU judgement). For products of dubious classification, such as certain new products, BTIs represent a practical solution to avoid disparities of treatment and ensuing disputes with customs authorities. However, since the tax classification is determined by the CN code, BTIs may also become a source of controversy between countries. Economic operators may be tempted to request a BTI in jurisdictions where it is more likely to obtain a more favourable (tax-wise) classification, in order to get competitive advantages across all EU national markets.

According to the database maintained by DG TAXUD, there are 1,025 alcoholic beverages in the EU that are covered by a BTI decision.53 Of course, BTIs do not only address products that intend to obtain a more favourable treatment, however a cursory analysis of the distribution across CN subheadings (Table 5) may provide a first hint of the areas where classification ambiguities prevail. In particular:

 Other fermented beverages, other than cider and perry (CN 2206 0039 and CN 2206 0059), alone account for a quarter of all BTIs, nearly as many as beer, wine and all other fermented beverages altogether.  There are also frequent BTIs in the area of aromatised wine product (AWP)54 below 18% vol (CN 2205 1010), which is another area of accelerated innovation.  The high concentration of BTIs in categories like other spirituous beverages (CN 2208 9069) and liqueurs and cordials (CN 2208 7010) may be partly explained by borderline products, e.g. certain ‘mixed drinks’ that did not manage to obtain a more favourable CN 2206 classification.

Table 5 – Distribution of BTIs across CN codes and EU countries

CN Heading and Sub-heading No. of BTI Most Frequent MS of emission All Beer (2203) 75 DE, UK All Wine (2204) 66 DE, FR Flavoured wine < 18% vol (2205 1010) 81 IT, FR

53 DG TAXUD, European Binding Tariff Information database, available at: http://ec.europa.eu/taxation_customs/dds2/ebti/ebti_consultation.jsp?Lang=en, last accessed on 02 May 2017. 54 In this Report, ‘aromatised wine products’ (abbreviated as AWP) refer generically to any kind of such products, including the three main subcategories laid down in Regulation (EU) No 251/2014: (1) aromatised wines; (2) aromatised wine-based drinks; and (3) aromatised wine-product cocktails (see Article 3 for the respective definitions). In practice, subcategory (1) is of little relevance for our analysis, and references to ‘AWP’ should be interpreted as primarily referred to subcategories (2) and (3) (sometimes made explicit in the text). It is important to highlight that our analysis focuses on fiscal classification of products and not to sectoral classification (as it is the case with Regulation 251/2014), so in our understanding ‘AWP’ includes all products that may be taxed accordingly, irrespectively of the ‘blurred’ boundaries between the product definitions established in Regulation 251/2014.

106

Rest of flavoured wine (2205) 15 FR, UK OFB, sparkling other (2206 0039) 89 UK, FI OFB, still other (2206 0059) 163 FR, DE, PT, UK, FI Rest of OFB (2206) 55 UK, FR Ethyl alcohol > 80% (2207) 61 NL, CZ Liqueurs and cordials (2208 7010) 64 IT, IE Other spirituous beverages (2208 9069) 258 FR, UK, CZ, SK, EE Rest of Ethyl alcohol < 80% (2208) 98 FR, FI, DE Source: Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s analysis of EBTI database (accessed in February 2017). Note: The CN codes reported in this Table relate to the latest version, i.e. Impl. Reg. 2016/1821.

Since the tax classification is determined by the CN code, BTIs may also become a source of controversy between countries and between economic operators. The following box highlights evidence of such disparities:

Box 2 – Review of selected BTI decisions (from the EBTI database)

Case 1 – Fruit-wine based alcoholic beverages with addition of ethyl alcohol

The UK classified as 2206.0059 (still other Poland classified as 2208.9069 (other spirituous fermented beverage other than cider and perry) a beverages) a mixture of fruit wine (obtained from mixture of cider (obtained from the fermentation of the fermentation of an unspecified fruit apple juice and sugar) and water, sugar, citric acid, concentrate and glucose syrup) and water, colours fruit flavours, colours and preservatives, fortified and flavours, fortified with the addition of ethyl with the addition of ethyl alcohol to bring the alcohol to bring the strength up to 21%. strength up to 21.9%.

Case 2 – Fruit-wine based alcoholic cream

The UK classified as 2206.0059 (still other Ireland classified as 2208.7010 (liqueurs and fermented beverage other than cider and perry) a cordials) a ‘country wine based cream’, with a ‘country cream’ obtained by mixing fermented 14.9% vol. apple wine with cream, with a 14.8% vol.

Case 3 – Wine-based ready-to-drink

Germany classified as 2206.0039 (sparkling other The Netherlands classified as 2208.7010 (liqueurs fermented beverage other than cider and perry) an and cordials) an aromatised wine-based drink aromatised wine-based drink made of: wine (white made of: wine, sparkling water, syrup, citric acid, wine or rosé wine), demineralised water, inverted and natural flavours, with an alcohol content of sugar syrup, citric acid, lactic acid, sodium 5%, due to the fact that – after the addition of benzoate, flavourings, colours, sulphur dioxide, water, sugar and flavourings – the ‘character of and carbon dioxide, with an alcohol content of 5% wine had been lost’. vol.

An aromatised wine product coded CN 2205 may be taxed according to three different categories, i.e. Art. 8, Art. 12 or Art. 17 depending on the addition of alcohol, the overall strength, and its specific denomination.

Pre-mix drink may be subject to Art. 12, Art. 17 or Art. 20 depending, again, on the alcohol origin and blend, the strength, and various country-specific rules.

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017

107

There have been cases where disparities in the interpretation of the CN classification prompted the issuance of a normative act that eventually repealed existing BTIs on certain products. In particular, this was the case with Regulation 1967/2005, which ruled that a certain beer product flavoured with tequila should be considered beer as defined under CN 2203.55

2 Excise duty classification

The excise duty classification is determined by the harmonised definitions laid down in Directive 92/83/EEC. According to Article 26 of the Directive, reference should be made to the CN ‘version’ in force at the time of adoption, i.e. Regulation 2587/91.56 However, as discussed above, the CN codes and the related explanatory notes are periodically revised. So Regulation 2587/91 is no longer in force, replaced by more recent ones (the latest being Commission Implementing Regulation 2016/1821). In this sense, the Directive contains references to CN codes that are, in principle, outdated. The issue was analysed in the Ramboll Evaluation, which concluded that it is not a source of practical problems and primarily a purely administrative issue (see Box 3 below). For this reason, and in agreement with the Commission, this matter is not in the scope of this Study.

Box 3 – Article 26 of the Directive and the issue of references to outdated CN codes

Article 26 establishes that the references to CN codes contained in the Directive relate to the version in force at the time of adoption, i.e. Regulation 2587/91. However, since Regulation 2587/91 is no longer in force, repealed by more recent version of the Combined Nomenclature (the latest being Commission Implementing Regulation 2016/1821), this means that the Directive contains references to CN codes that are, in principle, outdated.

On a closer look, the issue regards only two 8-digit CN codes no longer in use since recent versions of the nomenclature introduced further sub-headings. As shown in the excerpt reproduced in the Table in this box, there is no textual difference in the definition / description of the sub-heading in the two versions. The only difference lays in the fact that in Regulation 2016/1821 the numerical code is no longer mentioned. All other CN codes cited in the Directive are still valid today.

Table 6 – Comparison between outdated and updated versions of the CN codes

Commission Regulation 2587/91 Commission Implementing Reg. 2016/1821

55 Commission Regulation (EC) No 1967/2005 of 1 December 2005 concerning the classification of certain goods in the Combined Nomenclature, OJ L 316, 2.12.2005. 1. 56 COMMISSION REGULATION (EEC) NO 2587/91 OF 26 JULY 1991 AMENDING ANNEX I TO COUNCIL REGULATION (EEC) NO 2658/87 ON THE TARIFF AND STATISTICAL NOMENCLATURE AND ON THE COMMON CUSTOMS TARIFF , OJ NO L 259 OF 16. 9.1991

108

In theory, legal references to CN codes no longer in force may cause incongruences and uncertainties, but since the definitions have not changed there is no tangible consequences in using the outdated or the updated nomenclature version for the purpose of tax categorisation, in particular there is no risk that a product imported using a ‘new’ CN code could not be identified for excise duty purposes. Some CNEN have changed over time, but since CNEN are not legally binding (and are not explicitly mentioned in Article 26) they cannot fuel legal issues in the excise duty classification of the concerned products.

The matter was explicitly addressed in the Ramboll Evaluation, in particular:57

 According to Ramboll ‘these outdated references in the Directive were not reported by the stakeholders as a source of problems.’ (p. 116), and ‘there are no major, immediate and urgent negative consequences stemming from the reported inconsistencies’ (p.141). The results of our interviews confirm this conclusion.  Ramboll recommends to address this point in the next revision of Directive 92/83/EC. For the Commission this recommendation ‘concerns minor technical changes’ and is about ‘outdated references / good housekeeping’, i.e.: no relevant impact is envisaged (source: Commission Report to the Council COM(2016) 676 final). Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages

The five categories established in the Directive (see Table 7) are defined primarily with reference to CN 4-digits headings but the classification structure is partly different. In particular, there is no separate tax category for vermouth and other flavoured wine (like CN 2205) and there is only one category for ethyl alcohol including spirits instead of two – CN 2207 and CN 2208. On the other hand, the excise duty classification contains the ‘intermediate products’ (IP) category that is not present in the CN classification. The tax and customs classifications differ also at a more granular level, and the result is that each tax category of Directive 92/83/EEC may comprise of products that fall under multiple CN headings (see

Table 8).

A second important difference is that, unlike the CN classification, all tax categories are defined with reference to an explicit minimum and maximum alcohol strength, beyond which a product may change category, regardless of its nature. For instance, any fermented or mixed beverage, including wine, beer and cider, above 22% vol is taxed as ‘ethyl alcohol’. The harmonised tax categories may also contain reference to specific characteristics of the products, although not systematically (e.g. reference to the entire fermented origin of the beverage, enrichment etc.).

Table 7 – The five excise duty categories of alcohol and alcoholic beverages (Directive) Category Definition Any product falling within CN code 2203 or any product containing a mixture of beer with non- Beer alcoholic drinks falling within CN code 2206, in either case with an actual alcoholic strength by (Article 2) volume exceeding 0.5% vol.

Still Wine: (Paragraph 1) All products falling within CN codes 2204 and 2205, except sparkling wine as defined in paragraph 2: — having an actual alcoholic strength by volume exceeding 1.2 % vol but not exceeding 15 % Wine vol, provided that the alcohol contained in the finished product is entirely of fermented origin, (Article 8) — having an actual alcoholic strength by volume exceeding 15 % vol and not exceeding 18 % vol provided they have been produced without any enrichment and that the alcohol contained in the finished product is entirely of fermented origin. Sparkling Wine (Paragraph 2) All products falling within CN codes 2204 10, 2204 21 10, 2204 29 10 and 2205:

57 Ramboll, 2016.

109

— are contained in bottles with ‘mushroom stoppers’ held in place by ties or fastenings, or they have an excess pressure due to carbon dioxide in solution of three bar or more, — have an actual alcoholic strength by volume exceeding 1.2 % vol but not exceeding 15 % vol, provided that the alcohol contained in the finished product is entirely of fermented origin.

Still OFB (Paragraph 1) All products falling within CN codes 2204 and 2205 but not mentioned in Article 8 above, and products falling within CN code 2206, except other sparkling fermented beverages as defined in point 2 of this Article and any product covered by Article 2: — having an actual alcoholic strength by volume exceeding 1.2 % vol but not exceeding 10 % vol, — having an actual alcoholic strength by volume exceeding 10% but not exceeding 15 % vol, Fermented provided that the alcohol contained in the product is entirely of fermented origin. Beverages other Sparkling OFB (Paragraph 2) than Wine and All products falling within CN code 2206 00 91 as well as products falling within CN codes Beer (OFB) 2204 10, 2204 21 10, 2204 29 10 and 2205 not mentioned in Article 8 above which: (Article 12) — are contained in bottles with ‘mushroom stoppers’ held in place by ties or fastenings, or they have an excess pressure due to carbon dioxide in solution of three bar or more, — have an actual alcoholic strength by volume exceeding 1.2 % vol, but not exceeding 13 % vol, — have an actual alcoholic strength by volume exceeding 13%, but not exceeding 15 % vol, provided that the alcohol contained in the product is entirely of fermented origin.

All products of an actual alcoholic strength by volume exceeding 1.2% vol, but not exceeding 22 % vol and falling within CN codes 2204, 2205 and 2206 but not covered by Articles 2, 8 and 12.

Intermediate 2. Notwithstanding the provisions of Article 12, Member States may treat as an intermediate Products (IP) product any still fermented beverage falling within the scope of Article 12 (1) which has an (Article 17) actual alcoholic strength exceeding 5.5 % vol and which is not entirely of fermented origin,

and any sparkling fermented beverage falling within the scope of Article 12 (2) which has an actual alcoholic strength exceeding 8.5 vol. and which is not entirely of fermented origin.

— all products with an actual alcoholic strength by volume exceeding 1,2% volume which fall within CN codes 2207 and 2208, even when those products form part of a product which falls Ethyl Alcohol within another chapter of the CN, (ET) (Article — products of CN codes 2204, 2205 and 2206 which have an actual alcoholic strength by 20) volume exceeding 22 % vol, — potable spirits containing products, whether in solution or not. Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s analysis of EBTI database (accessed in February 2017).

The misalignment between the CN and the excise duty classifications may cause a certain degree of complexity in the categorisation of certain products. For example, an aromatised wine product (AWP) coded CN 2205 may be taxed according to three different categories, i.e. Articles 8, 12 or 17 depending on the addition of alcohol, the overall strength, and its specific denomination. Similarly, a ‘mixed drink’ may be subject to Articles 12, 17 or 20 depending, again, on the alcohol origin and blend, the strength, and other factors.

For the purpose of movement and monitoring within the EU, excise goods are given a harmonised Excise Product Code (EPC). The EPC are based on the tax categories described above, but do not fully comply with them. In particular, in the EPC system wine and OFB are merged (the distinction between still and sparkling products is maintained), which may be a source of ambiguity (discussed in Section 2.1.5.3). Secondly, ethyl alcohol and spirits falling under Article 20 are split into four EPC, as follows:

 B000 - Beer;  W200 - Still wine and still fermented beverages other than wine and beer;  W300 - Sparkling wine and sparkling fermented beverages other than wine and beer;  I000 - Intermediate products;  S200 - Spirituous beverages;

110

 S300 - Ethyl alcohol;  S400 - Partially denatured alcohol;  S500 - Other products containing ethyl alcohol.

For the reasons described above, the correspondence between EPC and CN codes is ‘many-to- many’ i.e. there can be several CN codes for the same Excise Product Code or vice versa.58 Table 8 below summarises the possible correspondences (not exhaustive) between the two systems, highlighting the cases where multiple correspondences are possible. In most cases, the correspondence is straightforward, nonetheless it is interesting to note that beverages under CN 2206 00 39 (‘other sparkling OFB’) may correspond to five different EPC, and similar degree of ambiguity can be found with various other CN 2206 products and – to a lesser extent – CN 2205 and CN 2204 products.

Table 8 – The multiple correspondences between CN and EPC59

CN headings / subheadings Excise product codes B000 W200 W300 I000 S200 S300 S400 2203 X

2204 10 and X 2204 29 10 2204 21 06 - 2204 21 09 X X

2204 21 11 - 2204 21 84 and 2204 29 11 - 2204 29 84 and X

2204 30 2204 21 85 - 2204 21 91 and 2204 29 85 - 2204 29 91 and X X

2204 21 86 - 2204 21 91 2204 21 92 and X 2204 29 92 2204 21 93 - 2204 21 98 and X X X 2204 29 93 - 2204 29 98 2205 10 10 and X X X 2205 90 10 2205 10 90 and X X 2205 90 90 2206 00 10 and 2206 00 51 and X X X

2206 00 81 2206 00 31 X X X X

2206 00 39 X X X X X

2206 00 59 and X X X X 2206 00 89

58 DG TAXUD, ‘Functional Excise System Specifications (FESS)’, version 3.65-EN, 16.09.2014. 59 EPC S500 is not displayed since it refers to products that does not fall in the CN 22 Chapter’s Headings for alcoholic beverages and spirits that are relevant in this Study.

111

CN headings / subheadings Excise product codes B000 W200 W300 I000 S200 S300 S400 2207 10 X

2207 20 X

2208 20 - 2208 70 and X 2208 90 11 - 2208 90 78 2208 90 91 X

2208 90 99 X

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s analysis of FESS-Appendix B.Note: CN headings / subheadings as in Regulation 2016/1821.

Finally, it is worth highlighting that the EU-level tax classification rules are sometimes complemented by national-level rules. These may regard the establishment of non-harmonised taxes for specific categories of products like the so-called ‘pre-mix’ or ‘alcopop’ tax in FR and DE; or MS-level distinctions within harmonised categories, such as the Romanian differentiation between cider & perry and other OFBs; or additional levies for products above a certain strength, etc. These specificities are based on domestic definition and criteria that add up to the harmonised ones and may create additional fiscal sub-categories that are relevant only locally.

112

ANNEX 8. DRIVERS OF THE DYSFUNCTIONS IN THE CLASSIFICATION OF CERTAIN ALCOHOLIC BEVERAGES

Driver 1: Lack of correlation between customs and fiscal classification

For legal certainty and to prevent the risk that the attribution of a certain classification is challenged (and fined) by customs or tax authorities when the product is already commercialised, economic operators may opt to apply for a BTI. These are classification decisions issued by the customs authorities of any Member State and they are binding throughout the EU for a period of normally 3 years. For products of dubious classification, such as certain new products or borderline products, BTIs represent a practical solution to avoid disparities of treatment and ensuing disputes with customs authorities. However, since the tax classification is determined by the CN code, BTIs may also become a source of controversy between countries and between economic operators. (See Annex 7)

Attempts to resolve excise classification issues through legal interpretation of the CN codes have not proved effective so far. Of the nearly 1 000 active BTIs for alcoholic beverages, approximately 50% of them concern products falling into 'borderline' product category. The database maintained by DG TAXUD provides an overview of the areas where classification ambiguities prevail. In particular:

 OFB, other than cider and perry (CN 2206 0039 and CN 2206 0059), alone account for a quarter of all BTIs, nearly as many as beer, wine and all OFB altogether.  There are also frequent BTIs in the area of aromatised wine product below 18% vol (CN 2205 1010), which is another area of accelerated innovation.  The high concentration of BTIs in categories like other spirituous beverages (CN 2208 9069) and liqueurs and cordials (CN 2208 7010) may be partly explained by borderline products, e.g. certain ‘mixed drinks’ that did not manage to obtain a more favourable CN 2206 classification.

BTIs apart, there are mismatches within the classification issue that relates to Excise Product Code (EPC). EPC are required for EMCS, the computerised system for monitoring the movement of excise goods under duty suspension in the EU. In the case of imports from third countries, the EMCS makes use of two distinct product classifications: (i) EPC and (ii) the customs CN code. EPC and CN codes have different origins and purposes, therefore the categorisations and the definitions used do not fully match. An EPC for each product category defined in the Directive is laid down in Regulation 684/200960. The correspondence between EPC and CN codes could be described as ‘many to many’ i.e.: (i) there can be several CN codes for the same EPC; or (ii) there can be several EPC for the same CN code61. In most cases, the correspondence is straightforward, nonetheless it is interesting to note that beverages under 2206 00 39 (‘other sparkling OFB’) may correspond to five different EPC, and similar situations can be found with all other 2206 products (OFB) and – to a lesser extent – some 2205 (e.g. vermouth) and 2204 (e.g. wine of fresh grapes) products. 'Still wine' and 'still fermented beverages' share the same EPC, which is also the case for sparkling wine and sparkling fermented beverages. The lack of a separate EPC for OFB is not ideal for market monitoring

60 Commission Regulation (EC) No 684/2009 of 24 July 2009 implementing Council Directive 2008/118/EC as regards the computerised procedures for the movement of excise goods under suspension of excise duty, OJ L 197, 29.7.2009, p. 24–64 61 DG TAXUD, “Functional Excise System Specifications (FESS)”, version 3.65-EN, 16.09.2014.

113

purposes or for MS which have a differential tax treatment. Furthermore the misclassification could result in the incorrect calculation of excise duty due and the financial guarantee required.

Driver 2: Unclear legislation for products manufactured using new technologies

Cider, perry, fruit wines, mead, etc. are agricultural products with a traditional origin, especially in Northern and Central Europe. These traditional fermented products were the target of Article 12 of the Directive as per CJEU jurisdiction. Article 12 is less strict than for other alcoholic beverages as this reflected the variety of national product practices for such beverages and the absence of a harmonised product definition and production rules that exist, for example, for wine and spirits.

The classification issues within the OFB category are essentially driven by the introduction of new production technologies and practices and the related development of products departing from the concept of ‘traditional fermented products’ for which the Article 12 and the category of OFB were conceived.

Overall, the problematic practices examined are of two main kinds:

(i) the use of an alcoholic base that has lost its essential fermented character

The fermented base used to produce an alcoholic beverage can be processed in various ways in order to obtain, among other things, the desired strength and a neutral or partly neutral organoleptic character. These are accepted processes - although with restrictions for certain types of beverages – that are intended to optimise and stabilise the taste and smell of the product, to compensate for the effects of weather and other crop-affecting events, as well as to innovate and develop products more in line with consumer expectations. These processes alter the fermented base through holding back or passing of some compounds in the beverage, allowing partial dehydration (concentration), partial dealcoholisation, tartaric stabilization, the adjustment of acidity and pH, reduction of the concentration of certain organic acids; management of dissolved gas, etc. There are different types of techniques that are used alone or in combination, in the production of a vast range of fermented beverages. There is no standardised description of these novel techniques that are at the moment, in most cases, only generically defined. In this sense, they lend themselves poorly to become subject to any regulatory provisions.

(ii) the addition of alcohol of distilled origin and other additives to a fermented beverage.

The addition of distilled alcohol is a well-established practice for several types of special wines and other traditional fermented beverages, and as such is regulated in sectoral legislation. However, alcohol is also added to mixed drinks with a fermented base to increase their strength. The economic rationale is that alcohol from distillation is generally cheaper to produce than from fermentation, and the addition of alcohol achieves the desired final strength in an easier and more flexible way. Directive 92/83/EEC does not clarify the amount of alcohol of distilled origin that can be added to a fermented base before the tax category changes. Similarly, the CN 2206 heading admits products not entirely of fermented origin62, but the permitted amount is not

62 The explanatory notes and classification opinions adopted by the HS Committee relating to Heading 2206 states: ‘All these beverages may be either naturally sparkling or artificially charged with carbon dioxide. They remain classified under this heading even when fortified with added alcohol or when their alcohol content has been increased by further fermentation, provided that they retain the character of products classified under this heading.’

114

specified63, and the jurisprudence in this area (i.e. the above mentioned C-150/08 case) did not establish any straightforward criteria. As a result, national customs administrations adopted different approaches to the classification of these products, combining objective criteria such as the share of distilled alcohol in volumetric terms or in terms of its contribution to the final ABV, and the subjective criteria laid down in CNEN note 2206 00. To the extent the differential in the tax rates applied to Articles 12, 17 and 20 products is high, there remains an incentive for economic operators to exploit this ambiguity.

63 When goods are prima facie classifiable under two or more headings, the CN rules require that classification is effectuated as follows: ‘mixtures, composite goods consisting of different materials or made up of different components, and goods put up in sets for retail sale, (…), shall be classified as if they consisted of the material or component which gives them their essential character, in so far as this criterion is applicable’.

115

ANNEX 9. TAX, AFFORDABILITY AND CONSUMPTION

To estimate the nature and the magnitude of the adverse effects potentially caused by the above classification uncertainties, it is necessary to consider at first the dynamics of the alcoholic beverage markets, including both the supply and the demand side. Needless to say, the mechanisms underlying this market are highly complex and the dynamics vary across market segments – in terms of both ‘type’ of beverage and price segment – and across geographical markets (i.e. consumption habits and consumer preferences).

This annex reviews in particular: (i) the mechanism of substitution across products and its connection with price levels; (ii) the effects of excise duty and its variation on consumer prices; and (iii) the possible general correlation between tax level, affordability and demand/consumption. The Study examined these mechanisms through an econometric analysis based on a combination of IWSR market data with tax levels and revenue data (Excise Duty Tables), also integrating other variables drawn from Eurostat and WHO GISAH. The Study triangulated the results with the results of similar exercises from the economic literature, and with the qualitative assessments collected from the stakeholder consultation.

 Cross-product substitution. This is conventionally measured through the ‘cross-price elasticity’. When this variable is positive, products are substitutes and the increase of price in one category results in an increased consumption of another category. When values are negative, products are complementary and follow the same trends (possibly influenced by an external third factors). When the correlation is not statistically significant, the analysed products are probably independent of each other.  The latter is frequently the outcome that can be found in the literature that tried to examine the cross-price elasticity of alcoholic beverages, which generally returns inconclusive and statistically weak evidence (see Box 4). In practice, no clear and robust substitution effect induced by price variations can be observed. In fact, substitution can be more substantially driven by factors other than price, and connected to: socio-demographic and lifestyle changes, marketing strategies, awareness-raising campaigns, national regulatory frameworks on labelling, commercialisation, and drinking etc. The list of variables can be very long and differs across contexts, so eventually the economic research has progressively abandoned the econometric approach based on cross-price elasticity. Moreover, it has been observed that price levels of different categories of products are often positively correlated. When prices fluctuate coherently for all products and nonetheless the level of demand varies, this would further confirm that consumption patterns, including substitution across products, is prevalently determined by other factors.

Box 4 – Selected excerpts from the literature review on cross-price elasticity for alcoholic beverages

A first review of estimates of cross-price elasticities in alcoholic products conducted in early 2000s64 showed a wide range of estimates of different sign, implying disagreement on whether beer, wine and spirits are complements or substitutes, and stressed the importance of extraneous factors, such as changes in consumer tastes and preferences. Still, the report concluded that the balance of evidence suggests that the drinks are substitutes, although cross-elasticities estimates have to be regarded with caution.

More recently, Meng et al. (2014) attempted to estimate the cross-price elasticities of off- and on-trade

64 Custom Associates Ltd, ‘Study on competition between alcoholic drinks’, 2001.

116

beer, cider, wine, spirits and ready-to-drinks in the UK by applying a pseudo-panel approach to the cross-sectional data on private households’ expenditures.65 Only 6 out of 90 estimated cross-price elasticities were statistically significant and the suggested substitution and complementary relationships were very difficult to explain (46 estimates had a positive signs and 44 a negative one).

A new study using cross-sectional data from the 2013 Australian arm of the International Alcohol Control survey employed a Tobit model approach to estimate cross-price elasticities of 11 categories of beverage, comprising on- and off-premise separately for regular beer (full strength), low-mid strength beer, bottle wine, spirits and ready-to-drinks, and off-premise cask wine.66 A significantly, positive relationship was detected between the prices of off-premise beverages with demand for the same beverage on-premise, while the cross-price elasticities among different beverage categories provided again indecisive results: very few statistically significant estimates (8 out 100) and a mix of positive and negative signs (49 and 51, respectively).

Given the above challenges, other studies on excise duties on alcoholic beverages excluded cross-price effects, which were regarded of secondary importance to the own-price effect.67

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages

The Study tested the cross-price elasticity for ‘borderline’ categories of products, through a variant of the model used to estimate the own-elasticity of individual categories of product. In particular, Economisti Associati tried to estimate if and how mixed drinks, ‘borderline’ ciders, and ‘borderline’ IP showed clear substitution correlations with other products. In some cases, a very small complementary correlation was registered, suggesting that other factors (such as the effect of the economic crisis, the introduction of certain national regulations, the impact of information campaign etc.) might have simultaneously influenced the level of consumption of different product categories, regardless of price. Overall, the outcome was not statistically significant, so no substitution could be demonstrated. 68

The above considerations should not be interpreted as a denial of the substitutability of all alcoholic beverages. In fact, this assumption is inter alia confirmed by certain marketing strategies, which increasingly abandon the approach ‘per class of product’ to adopt the ‘per consumption occasion / modality’ perspective (aperitif, refreshment, RTD, etc.). These complex marketing mechanisms, and their degree of success in moving consumers from one type of beverage to another, are outside of scope of this Study. For the purpose of this analysis, the key message is that a certain level of cross-products substitution cannot be systematically predicted by a variation in price.

65 Meng, Y. et al., ‘Estimation of Own and Cross Price Elasticities of Alcohol Demand in the UK. A Pseudo-panel Approach Using the Living Costs and Food Survey 2001-2009’, Journal of Health Economics, Vol. 34, 2014. 66 Jiang H. J., Livingston M., Room R., Callinan S., ‘Price elasticity of on- and off-premises demand for alcoholic drinks: a Tobit analysis’, in Drug and alcohol dependence, 2016. 67 London Economics (May 2010), Study analysing possible changes in the minimum rates and structure of excise duties on alcoholic beverages. 68 The very high correlation between prices across categories of product, which inflates the standard errors in multiple regressions (the ‘multicollinearity’ issue), and the fact that the model leaves out several independent, explanatory variables (the so called, ‘omitted variable bias’) do not allow firm conclusions about cross-price elasticities to be made. In the statistical model with fixed-effects at the product level, which allows for the controlling of time-invariant unobserved heterogeneity (e.g. consumer preferences across products stable over time), ten out 16 cross-price elasticities were statistically significant. However, this model is weak in dealing with time-variant unobserved factors at the product level (e.g. changes in consumer tastes over time that are different across product categories) and the results might be biased. Indeed, the statistical significance disappears when the HAC standard errors (heteroscedasticity and auto-correlation robust standard errors) are used, which allows accounting for serially correlated errors likely due to the previous omitted factors (e.g. consumer preferences across product categories that slowly change over time).

117

 The impact of excise duty on demand. From the above it derives that excise duty – that is one of the determinants of price – cannot have a statistically-significant correlation with cross-product substitution. In other words, at a systemic level, Economisti Associati have not observed a clear relationship between the tax rate applied on the target classes of products and the level of consumption of competing products. This statement requires two important qualifications: (i) despite the lack of a general relationship, under specific circumstances the variation of excise duty level can still have profound market effects; and (ii) as the results of this model show, despite the lack of a robust estimate for the cross-price elasticity, the ‘own-elasticity’ of specific categories of products can be estimated with a certain degree of precision.

With regard to the first point, the introduction of the ‘alcopop’ tax in Germany is a classical example of how taxes can indeed have a profound impact on substitution. This case, described in more detail in Box 5, was evidently caused by the very high level to which the tax was set and the fact that other potentially competitive products (malt and wine-based mixed drinks) were not targeted. However, in various other circumstances a significant increase of the excise duty applied to a specific category did not necessarily lead consumers towards other products. For instance, in 2013, the excise duty on beer in France increased by 160%, but the volume of sales continued to grow and no relevant changes were observed in other product categories.69

Box 5 – Possible substitution effects induced by the introduction of the ‘alcopop tax’ in Germany

Useful insights on substitution effects between different alcoholic products can be drawn from the review of the consumption trend of alcoholic beverages in Germany between 2000 and 2007. In the first three years of years 2000s, mixed drinks grew in popularity and their consumption recorded an impressive growth (about 78% per year, on average), which partly offset the decline in the volumes consumed of beer and spirits.

After the introduction of the alcopop tax in July 2004, consumers and the market responded negatively, and a major decline in consumption was recorded – i.e. amounting to some 50% per year between 2004 and 2006. Looking at the trend in consumption of other beverages, it seems that some previous drinkers of mixed drinks switched to beer as indicated by the slowing down of its declining rate (see Figure 1 below).

The existence of a similar substitution effect has been confirmed by a study conducted in 2010 to assess the effects of the alcopops tax on alcohol consumption and beverage preference among adolescents in Germany.70 Based on 2003 and 2007 data from the cross-sectional survey of the European School Survey Project on Alcohol and other Drugs (ESPAD), the study confirmed a partial substitution of alcopops by spirits and beer among 12–17-year-olds.

69 There have been changes in the excise duty levels of other products as well but very modest and on a much smaller scale than for beer. 70 Muller S, Piontek D, Pabst A, Baumeister SE, Kraus L., Changes in alcohol consumption and beverage preference among adolescents after the introduction of the alcopops tax in Germany. Addiction 2010; 105:1205–13.

118

Figure 4 - Indexed consumption of alcoholic beverages in Germany (in ‘000 hectolitres, 2001-2016)

110 300

105 250

100 200

95 150

90 100 drinks Mixed

85 50

80 0 2001 2002 2003 2004 2005 2006 Beer Spirits Wine All products Mixed drinks

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s elaborations on IWSR data. Note: For better readability the trend of mixed drink is displayed on a separate scale (right vertical axis).

Secondly, the results of this econometric analysis allowed us to estimate with sufficient degree of reliability the impact of excise duty variations on the demand of certain target categories of products. The exercise required two steps. The first step consisted in estimating the ‘pass-through’ effect of taxes on price, i.e. the average change of price level caused by a variation of excise duty rates (inclusive of the VAT on the excise duty). The impact on prices resulted more than proportional in the case of ‘borderline’ IP, cider, as well as various other non-target products;71 it is instead less than proportional for most of the IP products analysed and spirit-based mixed drinks.72 In the case of mixed drinks with a fermented base the relation observed is not statistically robust, i.e. it is not possible to predict the effect on price of a variation in the excise duty rate. This result is not surprising, given the generally short life-cycle of these products and the importance played by marketing strategies. So, for analytical purposes Economisti Associati assumed a conventional pass-through rate of 1 (i.e. a ‘full pass-through’).

The second step consisted in estimating the own-elasticity of the demand for the target categories of product, which in a nutshell is a measure of the variation of consumption expected when the price changes. Predictably, in all cases analysed the model returned negative coefficients, i.e. an increase in price would determine a reduction in the demand. Certain categories like ‘borderline’ IP and mixed drinks with a fermented base turned out very elastic, with the estimated drop in consumption much greater than the corresponding price variation. In some cases, the statistical robustness of the coefficient was lower, including for mixed drinks, so a certain variability exists in the reaction of consumers to

71 An increase by one EUR in the excise duty per litre has been estimated to translate into a change of the retail price per litre of EUR 1.33, EUR 1.73, and EUR 1.14 for ‘borderline’ IP, ‘borderline’ ciders, and various other non-target products, respectively. 72 The pass-through factor has been estimated at EUR 0.65 and EUR 0.28 for IP products, such as fortified wines and vermouths, and spirit-based mixed drinks, respectively.

119

price change, which can be again explained by exogenous factors like the impact of marketing and the volatility of these products. 73

 Tax, affordability and consumption. According to the World Health Organisation (WHO) database, the total alcohol per capita74 consumption in Europe has decreased by -10.4% from 2007 levels. This trend is confirmed by the decline in the sales of alcoholic beverages per capita in the EU that Economisti Associati estimated based on IWSR data.75 Accordingly, a decline of -4% was registered between 2010 and 2016, with an annual average reduction of about -0.7%. This reduction can be barely ascribed to a reduced average affordability of alcoholic beverages. The share of disposable income needed to purchase a fixed bundle of alcoholic beverages remained largely stable, recording a marginal increase from 1.73% in 2010 to 1.77% in 2015.76 The reason behind the stability of this ratio is that the average income growth in that period (+2.4% annually) largely kept pace with the average growth in the price level of alcoholic beverages (+2.8% annually).

Using a more accurate measure of the affordability, which considers how the price of alcohol has evolved as compared to the price of all other consumers’ goods, i.e. the Relative Alcohol Affordability Index (RAAI)77, the relationship between affordability and consumption is even weaker. As shown in Figure 5, the RAAI has risen over the last decade, driven by higher disposable income, whereas the indexed consumption declined, in an apparently unrelated manner.

73 Own-price elasticities for different groups of alcoholic beverages have been estimated by applying two common techniques for panel regressions, i.e. pooled ‘ordinary least squares’ (OLS) and ‘fixed effects’. Estimates achieved for different types of beverages, which have to be interpreted as the percent change in demand resulting from a 1% increase of their retail price, are the following: -1.15% and -2.99% for mixed drinks with a fermented base; (ii) -2.45% and - 2.47% for ‘borderline ciders’, (iii) -1.74% and -3.23% for ‘borderline’ IP, (iv) and -1.51% and -1.77% for other non- target products. A comparatively lower degree of statistical significance was found in the case of OFB groups of product; thus, in the case of ‘borderline’ ciders, the lower end of the range has been set at 1.3%, in line with the findings of the existing empirical literature (see, Stockwell, T.M. et al. (2012), ‘Does Minimum Pricing Reduce Alcohol Consumption? The Experience of a Canadian Province’, Addiction, Vol. 107, pp. 912-920; and Meng, Y. et al. (2014), ‘Estimation of Own and Cross Price Elasticities of Alcohol Demand in the UK. A Pseudo-panel Approach Using the Living Costs and Food Survey 2001-2009’, Journal of Health Economics, Vol. 34, pp. 96-103). 74 According to WHO, alcohol per capita (15+) consumption of pure alcohol is calculated as the sum of beverage-specific alcohol consumption of pure alcohol (beer, wine, spirits, other) from different sources. It is measured as litres of pure alcohol per person per year. 75 The per capita consumption is calculated as the ratio between the total volume of alcohol in litres consumed across Europe (from IWSR) for each category and the total national population (from Eurostat). 76 The bundle of alcoholic beverages is based on the per capita consumption of the five main categories of alcoholic beverages consumed in 2010 in EU, which included: (i) 70 litres of beer, (ii) 28 litres of wine, (iii) 5 litres of spirits, (iv) 0.6 litres of mixed drinks, and (v) 2 litres of cider. 77 We have used here the definition of the index provided by the UK National health Service (See: NHS Information Centre, ‘Statistics on Alcohol England, 2017 – Appendices’, the NHS Information Centre). The index has been recalculated at EU-level, based on the Eurostat’s harmonised indices of consumer prices and adjusted gross disposable income of households.

120

Figure 5 - Indexed trends in alcohol affordability and consumption (2005=100)

130

120

110

100

70 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

RAAI Alcohol Consumption

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s elaboration based on EBTI database (accessed in February 2017).

There are evident limitations to this analysis, namely: the fact that it looks only at broader systemic trends, it does not distinguish among specific socio-economic groups, and it does not distinguish specific ‘niches’ of products that might have become significantly more affordable (e.g. the renowned issue of ‘white cider’ in the UK).78 A micro-level perspective would be more informative in this respect, but at that level it is national/regional and not EU policies and measures that matter. So, for the purpose of estimating the impact attributable to Directive 92/83/EEC, the systemic-level analysis seems more pertinent.

The statistical analysis of the relationship between RAAI and per capita consumption suggests a positive relation, but with a small coefficient. In a nutshell, assuming all other factors neutral, a 1% decrease in alcohol consumption may require a 7% decrease in the affordability index. Under the strong assumptions that: (i) households’ disposable income grows at the same rate as the past 10 years (about 2% per year, on average), and (ii) the alcohol prices grow at the same pace of other consumables goods, such a leap in the affordability index would require an increase of the alcohol price by about 10%. Based on the IWSR data, the gross average price of alcoholic beverages in 2016 was about EUR 3.90 per litre, thus, a 10% increase will translate into an average increase in absolute terms of about EUR 0.40 per litre. With a conservative pass- through of excise duty on price equal to 100%79, and considering that the average excise duty levied on alcoholic beverages is EUR 0.68 per litre80, such effect on price would require a simultaneous increase of the excise duty rates by 57% (across all products in all MS).

78 https://www.theguardian.com/lifeandstyle/2011/apr/17/cider-industry-protected-expense-alcoholics 79 Our estimates of pass-through factor for different categories of beverages are typically higher than 1. These results are corroborated by the literature review. For instance, Sassi F. et al. (2103) conclude their meta-analysis of tax pass–through across different types of alcoholic beverages stating that: ‘[g]enerally, alcohol taxes are more than fully passed through to prices.’ (see, Sassi F., Belloni A., Capobianco C., ‘The Role of Fiscal Policies in Health Promotion’, OECD Health Working Paper No. 66, 2013). According to a recent study commissioned by DG SANTE pass-through coefficients appear more mixed across MS (see: Rabinovich, et al., ‘Further study on the affordability of alcoholic beverages in the EU - A focus on excise duty pass-through, on-and off-trade sales, price promotions and pricing regulations’, RAND Europe, 2012). 80 This value is based on the ratio between the total EU 28 revenue from excise duty on alcoholic beverage (EUR 35.6 bn) and the total volume of litres consumed (about 52 bn).

121

Historically, there are no known examples of comparable tax increases ever applied to the entire alcoholic beverage market, so there is no case-study evidence of the collateral effects of a similar fiscal measure. So, the above should be considered as a purely theoretical simulation whose main purpose is to show how moderate tax increases would only marginally affect overall consumption, at systemic level.81

The implication for the Study is that since the ‘borderline’ products have negligible impact on the affordability of alcohol (in systemic terms), they are unlikely to represent a huge threat to the general public health policy on alcohol control. Of course, at local level, certain ‘borderline’ products may constitute a social problem (e.g. too affordable or appealing to young people and vulnerable social categories), and in this sense they call for localised solutions, as has happened already in various MS.

81 In principle, the whole matter can be seen the other way round, i.e. a simultaneous, significant reduction of tax rates may induce an increase (or slower decrease) in the overall per capita alcohol consumption. This assumption is certainly compatible with the results of our Study as well as of other researches that investigated the link between taxation and demand (a relationship that is always mediated by the actual effects on retail prices). Some qualifications are nonetheless necessary: (i) the dependent variable in the above simulation is the consumption of ‘pure alcohol’, so the conclusions cannot be immediately extended to specific categories of product, especially products with a generally low ABV like beer, cider etc.; (ii) in particular, the reduced rates applicable in some MS to low-strength products may encourage the consumption of greater quantities, but this does not necessarily translate in an increase consumption of ‘pure alcohol’ (see Section 2.4 for details); (iii) the simulation regards a generalised variation in the price level, not only a segment of the market (as it is for instance the case with reduced rates for small producers, which – as discussed in Section 2.3 – concerns only a fraction of the overall EU market of alcoholic beverages).

122

ANNEX 9A EXCERPT FROM ANNEX 6A OF THE EXTERNAL EVALUATION (BY RAMBOLL)

4. EXAMPLES OF PRODUCTS DIFFICULT TO CLASSIFY

In this section, on the basis of our research into reported cases of products difficult to classify, we present the main groups (types) of products which cause the most difficulties. As relevant, we discuss the outcomes and root causes separately, for each group.

By far, the highest number of examples and the ones reported to have the most damaging consequences refer to products which are at the “border” between CN classification 2206 and 2208, meaning that from an excise perspective, they could fall in either” “Other fermented beverages”; “Intermediate products” or “Ethyl Alcohol”

4.1 READY-TO-DRINK PRODUCTS – LOW STRENGTH FERMENTED BEVERAGES OR MIXTURES (ALCO-POPS)

One of the most common products reported as being difficult to classify falls under the category of ready-to-drink products (also known as alcopops). This study has collected examples of such reported issues in at least 8 Member States82

In this category, products reported to be “difficult to classify” are normally between 4% and 7% ABV and consist of a fermented base with water, sugar, fruit juices, aromas and colorants.

The amount and proportion of alcohol coming from different origins (i.e. fermentation v distilled) is not material to define this category of products as difficult to classify because the examples consist of products with various shares of fermented to distilled alcohol.

Most examples provided were reported to contain alcohol of both fermented as well as distilled origin, although this was not a necessity. At least one example was reported to be based solely on alcohol of fermented origin which has been cleaned-up, in such a way as to lose its characteristics of a fermented alcohol

Common to these products is that in the opinion of those reporting the examples the products had lost the taste, smell and appearance of a beverage produced from a particular fruit or natural product and that it had the appearance and characteristics of a spirit drink, being labelled to that effect.

4.1.1 CLASSIFICATION

While in some Member States, they are considered to be (and taxed as) “Other Fermented Beverages” - (W200), sometimes even against the wishes and opinions of the respective tax administrations, in other Member States they are taxed as “Ethyl Alcohol” / “Spirituous Beverages” - (S200).

The excise classification described above follows the customs classification in the sense that products described above would be considered to be and taxed as “Other fermented beverages” in countries where they are classified for customs purposes as falling within CN 2206 and taxed as “spirits” in those countries where, for customs purposes they fall under CN 2208.

82 BE; DE; EE; FI; IE; PT; NL and UK

123

4.1.2 CONSEQUENCES

In all reported cases within this group of products, there is a clear difference between the excise taxes applicable when these particular products are classified as W200 as opposed to being classified as S200.

Specifically, depending on the individual variables of each case (i.e. the actual alcohol content of the product and the country of taxation), the difference in excise tax (VAT excluded) ranges from 7.48 EUR83 / HL to 89.7 EUR/HL84 of finished product.

In addition to the quantifiable difference in terms of applicable excise duty, economic operators interviewed have reported barriers to conducting business across the EU resulting from uncertainty with respect to the treatment of their product (i.e. being treated as W200/2206 in the home country, but considered S200/2208 in other Member States).

Finally, another negative consequence outlined by economic operators concerns competition aspects of the internal market. According to economic operators reporting examples of such products, the existence of this classification issue affects competition in two different ways:

 Firstly, it places producers of similar products (Ready-To-Drink products) which are based entirely on alcohol of distilled origin (which compete on the same market) at a severe competitive disadvantage (see point above on difference in taxation)  Secondly, it undermines the excise category itself (i.e. “Other Fermented Beverages”) by allowing RTDs to benefit from taxation at the same level as fermented beverages produced using traditional methods and natural fruits whose protection the category itself was supposed to benefit. Although not investigated within the scope of this study, the point of view of the consumer may also reveal a negative consequence if the expectations of consumers85 regarding these products are based on the assumption that they are based on spirits which have been pre-mixed to form a drinkable cocktail (as is the case with other categories of products difficult to classify86).

4.1.3 LEGISLATIVE SOURCE OF THE PROBLEM AND POTENTIAL SOLUTIONS

Due to the complexity of the problem and the similarities with products in other categories, this causes and solutions are treated separately in section 5.

4.2 MEDIUM STRENGTH FERMENTED BEVERAGES OR MIXTURES (10-15%)

Another group of products commonly reported as being difficult to classify is comprised of medium strength alcoholic beverages with a fermented base. Although similar in nature to the issue of “alco-pops”, the products falling in this group deserve a specific analysis because some legal considerations as well as potential outcomes surrounding them are rather different.

83 A ready to drink beverage of 5.5% alcohol in Estonia 84 A ready to drink beverage of 5.5% alcohol in Belgium 85 As investigating these aspects has fallen outside the scope of this study, this theory cannot be confirmed or infirmed in the context of this evaluation 86 See examples of Mixtures of Fermented and Distilled alcohol at approx. 21% ABV

124

Fewer examples of such products were reported to be difficult to classify, nevertheless they have been indicated as problematic products by the authorities of six Member States87.

In this category, products reported to be “difficult to classify” are between 10% and 15% ABV, with most of the examples being around 14%-15%.

Products in this group are slightly more diverse than “alco-pops”. However, they are all manufactured on a fermented base (either wine or other fruits), some of them being enriched with distilled alcohol.

Similarly to the ones described above, most examples provided were reported to contain alcohol of both fermented as well as distilled origin, although this was not a necessity. Common to these products is that in the opinion of those reporting the examples the products had lost the taste, smell and appearance of a beverage produced from a particular fruit or natural product and that they had the appearance and characteristics of a spirit drink, being labelled to that effect.

4.2.1 CLASSIFICATION

In the cases of these examples, there are actually three potentially applicable excise classifications: In situations where CN code 2206 is applicable the choice from an excise perspective would be between “Other Fermented Beverages” (W200) or “Intermediate products” (I000), although, in practice W200 would apply most often. If the product would fall for customs purposes within CN 2208 and “Ethyl Alcohol” (S200) excise classification would apply.

4.2.2 CONSEQUENCES

One of the reasons for grouping these examples as a different category than alco-pops was to showcase the difference in excise tax applicable to these products depending on interpretation of the provisions.

In all reported cases within this group of products, there is a large difference between the excise taxes applicable when these particular products are classified as W200 as opposed to being classified as I000 or S200.

Depending on the actual alcohol content of the product and the Member State where it is being sold, the difference in excise tax (VAT excluded) ranges from 79.55 Euros / HL (a 10-12% ABV, “Irish cream” type product in the UK) to 256.864 EUR/HL (a cleaned up fermented alcohol at 14-15% with sugar, aroma, acidifier, colouring and fizz in France) of finished product.

In addition to differences in terms of applicable excise duty, an important outcome reported in relation to these examples are litigation costs. Given the monetary impact at stake, disputes between tax administrations and operators in this area are more likely to be taken to court resulting in significant costs for both the administration as well as for the economic operators.

Although other negative consequences were not specifically mentioned by the stakeholders which have reported these examples, an adverse impact on fair competition could exist, should these types of products be in direct competitors (or be marketed to be in direct competition) with higher taxed spirits (e.g. those falling under I000 and S200).

87 DE; FR; HR; NL; PT; UK

125

4.2.3 EXAMPLES AND VOLUMES CONCERNED

One particularly illustrative example of products described above can be found in ECJ case C- 532/14 which concerns the excise tariff rate that is to be applied to alcoholic beverages that are based on fermented alcohol, known as Ferm fruit, to which distilled alcohol, sugar (syrup), milk, fats and various aromas are added. The alcohol percentage is in total 13.4%. At least 51% of the alcohol consists of fermented alcohol.

The fermented alcohol is cleared by means of ultrafiltration and has therefore a neutral taste, color and smell. The Tax Court88 considered the beverage as a liqueur, to be classified under CN code 2208 7010, upon which the high excise rate of distilled alcohol is due.

4.2.4 LEGISLATIVE SOURCE OF THE PROBLEM AND POTENTIAL SOLUTIONS

Due to the complexity of the problem and the similarities with products in other categories, this causes and solutions are treated separately in section 5

4.3 HIGH STRENGTH FERMENTED BEVERAGES OR MIXTURES (15-22%)

The highest number of products reported by stakeholders to be “difficult to classify” are beverages based on fermented alcohol which has been subject to certain production processes (e.g. ultra-filtration) or mixtures of alcohol below 21.9% to become colourless and odourless. Most often, this alcohol base mixed with other flavours (or without) is then marketed as a low- strength spirit.

Products sharing these characteristics have been reported in at least six Member States as well as by numerous stakeholders in the context of this case study.

The description of this kind of products, alongside the legal considerations and market distortions they cause is well illustrated by existing case law89.

4.3.1 CLASSIFICATION

In situations where CN code 2206 is applicable from an excise perspective, these products would be categorised as ”Intermediate products” (I000). If the product would fall for customs purposes within CN 2208 an “Ethyl Alcohol” (S200) excise classification would apply.

4.3.2 CONSEQUENCES

In all reported cases within this group of products, there is a very large difference between the excise taxes applicable when these particular products are classified as I000 as opposed to being classified as S200.

Assuming an alcoholic strength of around 21-22% and depending on the Member State where it is being sold, the difference in excise tax (VAT excluded) ranges from 200.00/ HL (a 21% ABV, fermented beverage in PT) to 331.40 EUR/HL (a 22% special fermentation of 'made wine' decolourised and flavour stripped then sold in Vodka style packaging in the UK) of finished product.

88 n.b. In the Netherlands 89 e.g. ECJ cases: C-150/08; C-532-14; C-533-14; UK case EWHC 17 (Ch) Diageo North America, Inc & Anor v Intercontinental Brands; etc.

126

A court decision involving one of the products described within this section90 recognised that the purpose of the product, at 22% was to benefit from lower taxation91.

Additionally, the impact on competition has been highlighted by the case law as well as by other stakeholders interviewed in the context of the case study:

 Firstly, the erosion of distinctiveness of the higher strength spirit92 which the products described in this category seek to imitate creates a quantifiable loss for the producers of the drinks being taxed as 2208/S200.  Secondly, the aspect of confusion of the consumers would further damage the legitimate interests of the producers of the alcohols being taxed as 2208/S200. Litigation results in significant costs for economic operators which will seek to correct the perceived unfair competition by these products.

Similarly to the above categories, although not investigated within the scope of this study, the point of view of the consumer may also reveal a negative consequence if the expectations of consumers regarding these products are based on the assumption that they are based on distilled rather than fermented alcohol93.

4.3.3 EXAMPLES AND VOLUMES CONCERNED

Below are just a few additional examples of these types of products sampled from Member States:

Croatia: An alcoholic drink with cherry flavour in a glass bottle of 500 ml obtained by the fermentation of apple juice which produced 13.05% vol. and enriched with distilled ethyl alcohol to the final alcohol content of 21%vol.

Ireland: Different brands which are marketed to appear almost as whiskeys, having an alcohol concentration of 22%, often made of cleaned-up fermented alcohol base to which distilled alcohol has been added.

Ireland: A golden-brown alcohol beverage at 21.9% manufactured using wine, sugar and/or flavours. The apple wine corresponds to 51% bulk volume and 61% of alcohol content.

United Kingdom: a 22% ABV product in a red get-up reminiscent of vodka. The front label includes the words "Premium" and "Imperial Blend". The back label states in relatively small print that the product is "a versatile blend of premium fermented alcohol with vodka.

Poland: Fermented beverages which have undergone filtration, fortified with distilled alcohol, to which flavours to change or strengthen smell or taste of the product have been added.

Netherlands: ECJ case C-533/14 concerns a beverage called Ferm Fruit (the base drink) with an alcohol percentage of 16%. This beverage is prepared with sugar syrup, demineralized water, apple concentrate, minerals and vitamins. After mixing, pasteurization takes place and wine yeast is added, as a result of which, the product becomes an alcoholic product. The alcoholic

90 England and Wales High Court (Chancery Division) 91 “Due to a favourable customs classification, a 22% ABV drink of this type would attract much less duty than spirits” 92 E.g. Vodka, Whiskey, Rum, Gin, Advokaat, etc. 93 As investigating these aspects has fallen outside the scope of this study, this theory cannot be confirmed nor infirmed in the context of this evaluation.

127

product is cleared by means of, among other things, ultrafiltration and has, therefore, a neutral taste, color and smell. It does not contain distilled alcohol.

4.3.4 LEGISLATIVE SOURCE OF THE PROBLEM AND POTENTIAL SOLUTIONS

Due to the complexity of the problem and the similarities with products in other categories, this causes and solutions are treated separately in section 5.

4.4 BEER TO WHICH ALCOHOL FROM DISTILLED ORIGIN IS ADDED

Mixtures of beer and spirits with recognisable increase in alcoholic strength: This type of product has been identified as difficult to classify by two stakeholders: One example was provided by authorities in Portugal while another example was provided by an economic operator in the UK.

In both cases, the beer had not lost its character as beer, but had rather gained a distinctive flavour of the aromatic substance spirit added (Tequila and Whiskey respectively).

The first example is a beer at 5.9% containing “water, malted barley, glucose syrup, corn, sugar, aromatic compounds (75% Tequila), citric acid, hop extract” while the second product is a beer, at 8% ABV, which has been matured for 12 months in a cask which previously had single malt whiskey stored within it. The beer grows in ABV, but is then diluted with beer until it reaches 8.1% ABV.

4.4.1 CLASSIFICATION

The excise codes applicable in this situation are CN2203 and CN2208 while the excise classification is, accordingly between “Beer” and “Ethyl alcohol”.

4.4.2 CONSEQUENCES

The difference of excise duty applicable depending on whether the product is classified as beer or ethyl alcohol is evident: in the two examples, it would range between approx. 52 EUR/HL (in PT)18 and 279.18 EUR/HL (in the UK).

18 The estimation of financial risk is not accurate in this case, as it was not possible to calculate the excise duty which would be applicable to the product in question (a beer at 5.9% alcohol) because the excise duty on beer in Portugal is expressed according to degrees Plato. As such, this figure is estimation.

The consequences of classifying such products go beyond the financial impact. In the case of the beer matured in whiskey casks, the tax warehouse in question (as well as some of its customers) had to adapt their authorisations as they were not allowed to hold / receive S200 products under suspension of excise duty, leading to high administrative costs and unexpected liability to pay tax resulting from releasing for consumption of the product in question.

4.4.3 LEGISLATIVE SOURCE OF THE PROBLEM AND POTENTIAL SOLUTIONS

Article 2 of the Directive only foresees products of 2203 to be beer (or “any products containing a mixture of beer with non-alcoholic beverages falling under CN2206”).

In the examples provided, it has been argued that CN code 2208, as it is defined, is able to capture, as ethyl alcohol, a beer to which only an 0.1% ABV can be attributed to alcohol of distilled origin, even though the vast majority of alcohol is sourced from the fermentation of malt.

128

While this study has absolutely no authority to judge whether a given (or claimed) classification is correct or not, from the examples provided in the context of this study, we believe that the current legislative framework is sufficient to provide an accurate determination of the products in question, and therefore we see no systematic weakness of the Directive as regards to this issue which would necessitate revision.

Below, we outline the legal considerations which we would deem as sufficient for an unequivocal judgement in these cases.

As we recall the general rules for the interpretation of the CN (‘the general rules’), which appear in Part One, Section I A, of the CN, provide inter alia:

‘Classification of goods in the [CN] shall be governed by the following principles:

1. …

2. …

(b) Any reference in a heading to a material or substance shall be taken to include a reference to mixtures or combinations of that material or substance with other materials or substances. … The classification of goods consisting of more than one material … shall be according to the principles of rule 3.

3. When by application of rule 2(b) or for any other reason, goods are prima facie classifiable under two or more headings, classification shall be effected as follows:

(a) The heading which provides the most specific description shall be preferred to headings providing a more general description. However, when two or more headings each refer to part only of the materials or substances contained in mixed … goods …, those headings are to be regarded as equally specific in relation to those goods, even if one of them gives a more complete or precise description of the goods;

(b) Mixtures … which cannot be classified by reference to 3(a), shall be classified as if they consisted of the material … which gives them their essential character in so far as this criterion is applicable.

In our assessment, and without prejudice to any decisions of the Member States and of the courts, we believe that the current legislative environment should be sufficient to determine the unequivocal classification of the products in question.

4.5 WINE BASED DRINKS (FLAVOURED WINES AND APERITIFS)

Several examples of wine based drinks were reported to be “difficult to classify”, highlighting two distinct issues:

1. The first group of examples referred to wine based drinks to which flavours containing alcohol of distilled origin is added. According to the respondents reporting these products, the characteristics of these examples remain that of wine, and alcohol content is maximum 14%. - Usually, the added alcohol (from distilled origin) content is between 0.5% and 1.2%.

2. Another category of products refers to flavoured wines, either mixed with fruit aromas, subject to crio-extraction but remaining below 15%.

4.5.1 CLASSIFICATION

129

In both cases the products described will remain for customs purposes either CN 2204 or CN 2205.

However, the problems highlighted for the two types of products are different:

1. While for the first category (of wine flavoured with the addition of distilled alcohol, but below 15%), some countries would consider the same product, for excise purposes as W200 (“Other fermented beverages”) while others would consider it as I000 (“Intermediate beverages”).

2. In the second category (those of flavoured wines, to which no distilled alcohol is added), they remain in the excise category of W200 (“Other fermented beverages”).

4.5.2 CONSEQUENCES

1. This particular example does not concern a difficulty to classify a product within one country, but the treatment of the same product in different Member States. As a result, this is mainly an internal market issue; the same producer has the product classified differently in different MS.

2. In the second example, competitive distortions as a result of differing excise rates for products being argued to compete on the same market have been reported to be the most important consequence. This relates to the treatment of certain products in France, these products would arguably be perceived to be vermouths, competing on the market of aperitifs but classified as W200, in the detriment of competing products which are being considered to be I000. In this case, the financial difference, in terms of excise duty is approximately 185 EUR/HL (3,77 € / hl vs 188,41 €/hl).

4.5.3 EXAMPLES AND VOLUMES CONCERNED

The example concerned in the second scenario presented in this section refers to the treatment of a well-known and popular brand of vermouth in France

4.5.4 LEGISLATIVE SOURCE OF THE PROBLEM AND POTENTIAL SOLUTIONS

1. In the first set of examples, the issue at stake was that Art 12 (1) - the concept of entirely fermented origin is being interpreted differently. Furthermore - Art 17 (2) - where “Intermediate products” fall is interpreted differently by different Member States

It was reported that, in some Member States, the addition of flavours with alcoholic content to a wine base product is possible without the loss of the excise classification as W200. However, national legislation is not harmonized in this respect, leading to situations where the same product is classified as an intermediate product in some Member States and as a fermented beverage in others. As an example, it was reported that in Spain, an addition of distilled alcohol to wine of 0.5% would be permitted without changing the classification for excise purposes, while in Italy, a threshold of 1.2% is applicable. In this situation, a product which may be “Other Fermented beverages” in Italy, will be an “Intermediate product” in Spain.

The Directive should be clear to define the situation of adding flavours containing alcohol to wine. In order to have uniformity in deciding at which point the product becomes an intermediate product. To this respect the Directive should seek to clarify the notion of “entirely of fermented origin” within the understanding of Articles 8, 12(1) and 17. It should be noted that the problems reported in this section are different in their scope and nature than those reported under the sections above, and a solution to one may have unintended consequences on the evolution of the problems in the other category.

130

2. The legislative source of the issue in the second example is the application of excise legislation which classifies the products in question19 as such in spite of arguably sharing similar characteristics and competing on the same market. No solution to this particular issue can be found within the current legislative context.

4.6 OTHER ISSUES

Finally, an inconsistency of the Directive has been reported in the context of this evaluation, it refers to the manner in which sparkling wine is defined for excise and customs purposes: Sparkling wine for excise purposes, defined in Article 8 (2) indent 1 requires a pressure of 3 bar or more while the equivalent CN codes require more than 2.5 bar. This mismatch requires more than 2.5 bar. This mismatch between classifications should be resolved.

131

ANNEX 10. REDUCED RATES FOR OTHER SMALL PRODUCERS

IMPLEMENTATION OF REDUCED RATES FOR SMALL BREWERIES AND DISTILLERIES ACROSS MS

The majority of MS – 23 out of 28 – have opted in to the reduced rates for small breweries. Thirteen out of these 23 MS have adopted the maximum threshold allowed by the Directive, the remaining 10 a lower one, from as low as 6 000 hl/year in Estonia, up to 150 000 hl/year in Finland. Eleven MS have established a bracket system, with two to five brackets, i.e. they provide a larger discount for very small breweries compared to the one granted to those whose output is close to the threshold. While most of the MS provide for a fixed discount rate (for each bracket where applied) expressed in EUR per hectolitres / Plato degree or EUR per hectolitres / ABV, three MS – Denmark, Poland, and the United Kingdom – have a slightly more complex system where the discount decreases proportionately as the output increases. Not all MS provide the maximum allowed discount – i.e. 50% of the normal rate – or they provide the full discount only for the smallest output bracket. Information is summarised in Figure 3 below.

Figure 3 – MS implementation of the reduced rates for small breweries (2017)

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", EDT. Legend: In blue: MS without reduced rates for small brewers; in white: MS with reduced rates for small brewers and maximum threshold; in grey: non-EU countries. Numbers indicate the threshold applied. Note: (*): degressive system.

Analysing how MS have implemented this provision, the following considerations emerge:

 There does not appear to be an inverse relation between the excise rate level and the

132

decision to grant reduced rates. For instance, Spain and Sweden are among the MS with a higher rate on beer, but they did not opt in to the provision, while low-excise MS such as Bulgaria, Germany, Romania, or Latvia did opt in.  Smaller MS tend to have lower maximum thresholds, but this is not always the case. For example, small-to-medium MS, such as Belgium, Denmark, Malta, Portugal, or Luxembourg, did adopt the 200 000 hl per year limit, while large MS such as Germany or the UK grant no reduction above 40 000 and 60 000 hl per year, respectively. This results in a very different market share potentially covered by a small brewery, from as low as 0.05% in Germany to as high as 67% in Luxembourg or 92% in Malta.  A microbrewery producing 1 000 hl per year receives the maximum possible reduction (50%) in 14 MS out of the 23 opting in to the provision, while in 3 MS it receives a limited reduction (less than 15% of the standard rate).  A small brewery producing 10 000 hl per year is granted the maximum possible reduction (50%) in 10 MS out of the 23 opting in to the provision, while in 5 MS it enjoys no or limited reduction (less than 15% of the standard rate).  A medium brewery producing 100 000 hl per year is granted the maximum reduction (50%) in only 5 MS out of the 23 opting in to the provision, while in 12 MS it enjoys no or limited reduction (less than 15% of the standard rate).

Reduced rates for small distilleries have a much lower implementation rate, as only 7 MS have decided to apply it: Austria, Germany, Spain, Croatia, Portugal, Romania, and Slovenia. In Slovenia, the yearly output is set at 1.5 hl of spirits per year; in Austria, the yearly output threshold is 4 hlpa per year, while, in the other 5 MS, the output threshold corresponds to the maximum allowed by the Directive (10 hlpa). All MS provide for the maximum possible discount (50%), except for Austria and Germany, which come close to it (46% and 44% respectively), while in Spain the discount amounts to 12% of the standard rate. There are no output brackets or decreasing reductions. Details are provided in Figure 4 below.

133

Figure 4 – MS implementation of the reduced rates for small distilleries (2017)

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", EDT. Legend: In blue: MS without reduced rates for small distilleries; in white: MS with reduced rates for small distilleries and maximum threshold; in grey: non-EU countries. Numbers indicate the threshold applied. Note: (*): threshold expressed as hl of spirits.

NATIONAL SCHEMES FOR SMALL PRODUCERS OF ALCOHOLIC BEVERAGES

In the six MS visited for this policy issue, two additional schemes providing reduced rates or full exemption to small producers have been identified: the Abfindung for small distilleries in Austria and the exemption for small cider makers in the UK.

 The Abfindung scheme for small distilleries in Austria.94 In Austria, farmers traditionally distil their own fruit production, for their own consumption as well as for sale. Though its economic importance in terms of farmers’ income is considered to be marginal, distillation is regarded as part of the Austrian rural culture. Rules for farmers’ distilleries have been in place for more than 250 years. In Austria, two kinds of distilleries may be set up:

o Verschlussbrennerei (sealed distillery), in which the duty to pay is calculated on the exact amount of alcohol produced. o Abfindungsbrennerei (small-scale flat-rate distillery), in which the excise duty is calculated on an estimated output.

The Abfindung regime is defined in Article §55 of the Austrian alcohol tax law95 An

94 A similar scheme exists in Germany as well, as detailed in §114 of BrennO 1998. 95 ‘Bundesgesetz über eine Verbrauchsteuer auf Alkohol und alkoholhaltige Waren (Alkoholsteuergesetz)’, consolidated version of 21.04.2017.

134

Abfindungsbrennerei can produce up to 2 hlpa per year; on the first hlpa, a reduced rate equal to 54% of the standard rate is applied; on the second hlpa, a reduced rate equal to 90% of the standard rate is applied (§65). Products from an Abfindungsbrennerei can be put up for sale under certain restrictions, but cannot be sold in other EU MS (§57). Any individual can apply to carry out distillation activities under the Abfindung regime, by registering as a producer and communicating to the customs authority his/her distilling equipment, the raw materials that will be used, and the timing and duration of the distillation. According to the raw materials used and the duration of the distillation, an output is estimated, and the excise duties are calculated. Only own fruit or other agricultural products can be distilled in an Abfindungsbrennerei.

 The small cider maker exemption in the UK. The United Kingdom has an exemption from excise taxes for small cider makers producing less than 70 hl per year. The exemption dates back to 1976, and was contextual to the introduction of excise duties on cider. The UK government announced in July 2015 that it would retain the exemption ‘until and unless a replacement scheme is established’.

To be eligible for the exemption, small cider makers must apply for an authorisation from the customs authority. Once the authorisation is obtained, small cider makers are exempt from the various excise obligations (e.g. recordkeeping, auditing, excise payments, setting up of a tax warehouse). The customs authority performs occasional checks, and further investigates if anything appears suspicious.

OTHER PROVISIONS FOR SMALL PRODUCERS

Though reduced rates are not granted to small producers of wine and other fermented beverages, these may be granted an exemption from most of the administrative requirements provided by the excise legal framework. Article 40 of the Horizontal Directive provides for MS the possibility to exempt small wine producers from the requirements on (i) production, processing and holding (including the setting up of a tax warehouse);96 (ii) movement of excise goods under suspension;97 and (iii) any other requirement relating to movement and holding. Small wine producers are defined as those with an output of less than 1 000 hl of wine per year. Based on Article 15 of Directive 92/83/EEC, this provision can also be applied to other fermented beverages.98 According to the fieldwork carried out and to the recent evaluation of the Horizontal Directive, Austria99 and Italy100 apply the exemption; on the contrary, France does not, but it exempts small winegrowers from lodging an excise guarantee.101

Industry analysis

In the following sub-sections, data on the various beverage industries are presented, in order to estimate the number of small players and their market share. For each tax category included in

96 Chapter III of the Horizontal Directive 97 Chapter IV of the Horizontal Directive 98 As done e.g. by Italy, see Article 8 of ‘Decreto 27 marzo 2001, n. 153, Regolamento recante disposizioni per il controllo della fabbricazione, trasformazione, circolazione e deposito dell'alcole etilico e delle bevande alcoliche, sottoposti al regime delle accise, nonché' per l'effettuazione della vigilanza fiscale sugli alcoli metilico, propilico ed isopropilico e sulle materie prime alcoligene’, consolidated version of 4.7.2017. 99 ‘Schaumweinsteuergesetz 1995’, Part 3, §44 (3). 100 See Article 37.1 of ‘Decreto Legislativo 26 ottobre 1995, n. 504, Testo unico delle disposizioni legislative concernenti le imposte sulla produzione e sui consumi e relative sanzioni penali e amministrative’, consolidated version of 4.7.2017. Hereinafter: ‘Italian Excise Law’. 101 Code général des impôts, art. 110-D.

135

the Directive, the most representative product is analysed: beer, still wine, cider for the other fermented beverages, distilled spirits for ethyl alcohol, and fortified wine for intermediate products. These industries produce the most common products in their category in the six sample MS.102 Indeed, the supply and market share analysis presented below needs to rely on a relevant market, defined along geographical boundaries (i.e. for each sample MS) and product boundaries. For the definition of the policy options and the impact analysis (in Section 3.3 below), consideration will be given to the extension of reduced rates to the whole tax category.

THE BEER INDUSTRY

Five of the six sample MS apply reduced rates for small brewers. France, Belgium and Poland grant them up to an output of 200 000 hl per year, while Austria and the United Kingdom limit it respectively to 50 000 and 60 000 hl All MS but France provide for a bracket system, the discount being higher for smaller entities. A reduction up to 50% of the standard rate is granted only by France and the UK, while all other countries provide for a lower discount.103 Importantly, when breweries get close to the threshold, the reduced rate gets closer to the standard one, at 90% or more of the latter, hence the tax advantage becomes smaller. Full information is reported in Table 9 below.

Table 9 – Implementation of reduced rates for small brewers in the sample MS Output upper limit Brackets Reduced rate MS Standard Rate (hl) (hl) (% of standard) 0-12,500 60% 12,500-25,000 70% AT 2.00 €/hl/° Plato 50,000 25,000-37,500 80% 37,500-50,000 90% 0-12,500 87% 12,500-25,000 90% BE 2.00 €/hl/° Plato 200,000 25,000-50,000 93% 50,000-75,000 96% 75,000-200,000 99% 7.41 FR 200,000 No as of 2013 50% €/hl/% vol 0-20,000 68%* 20,000-70,000 84%* PL 1.81 €/hl/° Plato 200,000 70,000-150,000 87%* 150,000-200,000 90%* 0-5,000 50% 21.04 UK** 60,000 5,000-30,000 86%*** €/hl/% vol 30,000-60,000 97%*** Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages, EDT. Notes: * calculated on 12° Plato beer; ** with reference to beer with 2.8-7.5% vol; *** calculated on the mid-point of the bracket.

Table 10 below provides a summary of the information retrieved on the production structure of the beer industry in the six MS. These best estimates result from the consolidation of public data

102 Based on ISWR sales data for 2016 and total excise revenues per fiscal category from EDT. 103 In Poland, the reduced rate is granted per hl of production, while the excise is calculated per hl/° Plato. Hence, the ratio of the reduced rate over the standard one is not fixed.

136

provided by tax authorities and trade associations, as well as from quantitative and qualitative information collected during the fieldwork and used to complement missing data. The definition of the various size classes varies from country to country, as there is neither a shared industry consensus, nor a standard data collection format. In most countries, the definition of what a micro or small brewer is depends on the national brackets used to administer the reduced rate scheme. The definition of a medium brewer is largely influenced by the country market structure.

Table 10 – Number of small brewers and supply structure AT BE FR IT PL UK EU* % of brewers covered by 0% the reduced rate 90% 97% 99% (98% below 87% 99% 97% 1000 hl) % of output covered by 6% 10% 4%. 0% 3% 5% 5% reduced rate Market Share Micro Brewers 1.5% (up to 1,000/5,000 hl) 3% 1.5% 2.5% 2% 5% Small Brewers 8.5% (up to 10,000/20,000 hl) n/a Medium Brewers 15% 7% 2.5% 6.5% 8% 20% (up to 100,000/200,000 hl) Large Brewers 75% 90% 96% 92% 90% 75% (over 200,000 hl) Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages, Brewers of Europe: interviews with national trade federations, tax and customs authorities. Note: market segment covered by the reduced rates; market segment partly covered by the reduced rates. (*): data refer only to MS implementing the reduced rates; based on values from sample MS (AT, BE, FR, PL, and UK). Size classes are only indicative and vary across the sample MS, depending on the thresholds adopted for reduced rates and industry practice.

Findings show that the vast majority of active brewers, 97% in the overall sample, and about or more than 90% in each country, is covered by the reduced rates. However, their production represents a small share of output, 5% in the overall sample, and not higher than 10% in any MS. In the UK and Austria, where the output threshold is lower than the maximum allowance and where the market features a significant group of so-called regional brewers in the area of 100 000 to 500 000 hl, reduced rates cover 5-6% of the market. In Belgium, a country with a longstanding tradition of local and small brewing and where the maximum threshold is set at 200 000 hl, reduced rates cover 10% of the market, the highest share among the MS analysed. In Poland, there is a significant presence of mid-size breweries, with an output of about 100 000-500 000 hl per year; however, since only a part of these regional players falls below the threshold, the market share covered by reduced rates is about 3%. In France, the importance of small players is limited: even though the number of small brewers increased four-fold over the last decade, their population has grown from a very small base;104 as a consequence, small brewers represent only about 4% of the market. Finally, in Italy, where reduced rates do not exist, microbreweries up to 1 000/1 500 hl represent 98% of the active players and about 2.5% of the market.105

THE DISTILLED SPIRIT INDUSTRY

The quality of the information on the number, size, and market share of players active in the

104 In France, the brewing tradition almost disappeared in the 1980’s, when only 30 brewers were active; this was due to changes in consumer taste and the industry consolidation process. This trend has reversed and, nowadays about 1,000 active operators exist. E. Gillard, ‘Bières et brasseries françaises du 21ème siècle’, Projet Amertume, 2016. 105 These findings are in line with those described in the Ramboll Evaluation, where the share of production covered by reduced rates in the three countries for which data are estimated is around or above 90%, and where the 5 big producers control 50% to 70% of the market. MS covered: DE, FR, IT, and UK.

137

supply of distilled spirits is much poorer compared to what is available for beer and wine.106 This is also due to the fact that only 7 MS opted in for the reduced rates for small distilleries – while 23 opted in for beer – and to the fact that there is no definition of ‘small spirit producer’ in other parts of the EU legislative framework, either for tax or agricultural policies. However, several trends emerged from the fieldwork, which can be summarised as follows:

1. The number of active distilleries is in the order of magnitude of 100 units in four out of the six MS visited: in particular, 120 distillers are licensed in Poland, between 75 and 90 of which are considered active; about 150 are active in Italy, and 230 in the UK. In Belgium, it is estimated that about 40-45 active distilleries are present in the market. In France, the number of operators is estimated at about 5 000-10 000. Austria is an exception: therein, it is estimated that about 30 000- 40 000 companies or individuals, mostly farmers, distil spirits – the vast majority under and because of the simplified flat-rate Abfindung regime.

2. Data on the size of distilleries are scant. In Poland about 45 distilleries produce less than 100 hlpa per year and their market share is estimated to fall below 0.4%. In France, 50 to 60 distilleries are estimated to fall below 10 hlpa, and they would represent, at maximum, 0.04% of the spirit market. To the contrary, estimates show that about 2 000 French distillers produce less than 10 000 hl of spirits per year (equivalent to 4 000 hlpa at 40% vol). In the UK and Italy, stakeholders and the authorities estimated that the presence of small distilleries with a scale of 10 hlpa is nihil or negligible, and that they could be active only in very premium segments, or as ancillary activities to farming, with a strict local dimension. In Austria on the contrary, most of active distillers fall within the Abfindung, and hence produce only up to 1 or 2 hlpa per year.

3. There are growth trends in the small distillery segment, but they are not widespread across the MS. Growth was reported both in the UK, also thanks to a spur of small gin distilleries, and in Belgium, based on data on the applications for a tax warehouse. Growth of small distillation is not driven by fiscal incentives, but rather by consumers’ demand. However, in other countries such as Poland, the number of distillers, and especially of small agricultural distilleries, is rapidly shrinking; in Italy, there is no indication of a growth of small-scale distillation.

THE CIDER INDUSTRY

The consumption of cider is largely concentrated in a handful of MS. The UK has, by far, the largest market, representing about 50-55% of the EU market, followed by Spain, France, Germany, and Ireland.107 Cider markets are larger in countries where there is a traditional production. The most important cider-producing countries or regions are indeed the British Islands – both Ireland and the UK – France, especially Normandy and Brittany, Spain, especially the Asturias, and Germany. To better collect information on cider, the sample of MS thus includes Ireland, while Austria and Belgium have been dropped. In this way, the sample consists of three of the largest cider markets (France, Ireland, and the UK) and two marginal producers (Italy and Poland).

Table 11 below provides a summary of the information retrieved on the production structure of the cider industry in these five MS. The estimates below are based on data provided by customs, tax, and agricultural authorities and trade associations. Additional quantitative and qualitative

106 As also acknowledged by trade associations. See Vinum & Spiritus Association Belgium, ‘La réalité économique derrière notre secteur’, Available at: http://fr.vinumetspiritus.be/sector/economics/de-economische-realiteit-achter-onze-sector/, last accessed on July 2017. 107 AICV, ‘European Cider Trends’, 2016 Update.

138

information collected during the fieldwork was used to complement missing data. The definition of the various size classes varies from country to country, as there is neither a shared industry consensus, nor a standard data collection format. In general, micro-cider makers are considered to be those whose yearly production is below 100 hl (e.g. 70 hl in the UK). For these players, cider production remains an ancillary activity, e.g. for farmers or farmhouses. The definition of small cider makers usually encompasses those whose production is below 10 000/15 000 hl.

Table 11 – Estimated market share of small and micro cider makers and their output

FR IE IT PL UK EU

% of small cider makers 10% 20% n/a 32% 15% 100% % of micro-cider 82% 89% 73% 11% 64% makers Production share of 17.5% 3% n/a 3% small cider makers 100% 4.6% Production share of 2.5% 0.1% 2% 0.1% micro-cider makers Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages, interviews with national trade federations, tax and customs authorities, ministries of agriculture. Note: EU estimates based on the five sample MS.

The distribution of the firm population and production of cider is similar to that of the beer industry. Micro- and small cider makers represent about 97% of the number of active companies, and between 93% and 99% in the MS considered. Their production share is estimated at 4.6% in the MS considered, and is below 5% in all countries except for France, where small independent companies are estimated to represent between 15% and 20% of the production. The production of micro-cider makers is negligible in the UK and Ireland, where the market is dominated by very large companies, and does not reach more than 3% in Poland and France, confirming the ‘ancillary’ nature of this market segment.

THE STILL WINE INDUSTRY

The still wine value chain features different actors that play different roles in terms of scope of the activity. There are at least four kinds of wine producers:

1. ‘Classical’ wine makers: companies that are both wine growers and wine makers; they produce and bottle their own wines. A wine maker may also buy grape, juice or bulk wine from other producers, in a variable percentage. 2. Independent wine makers: as the classical wine maker, they are both wine growers and wine makers. However, independent wine makers process only their own grape. They usually have a smaller scale than classical wine makers. 3. Cooperatives of wine growers: a cooperative collects grape, juice or bulk wine from its members, which in turn usually are co-owners, who then receive monetary or in-kind compensation as a share of profits or finished production. Cooperatives are thus wine makers, but not necessarily wine growers. Very small wine growers, who have no interest or no means to produce, bottle, and trade wine, usually confer their production to cooperatives. 4. Negociants en vin, or wine shippers: a wine shipper buys grape, juice, or bulk wine from wine growers, and then produces, bottles and sells wine under its own name. Hybrid companies

139

exist, which are wine makers and also produce wine as shippers.

As a consequence, there could be at least two kinds of small players in the still wine value chain: the small wine grower – regardless of whether it confers his/her production to a large player or not – and the small wine producer. As the focus of the reduced rate provisions is on operators producing alcoholic beverages (i.e. breweries and distilleries), rather than on other operators along the value chain, the analysis below focuses on small wine producers. However, it should be remembered that large wine makers often work in cooperation with a constellation of small players.

Table 12 below provides information on the share of players below 1 000 hl in the six sample MS, and on their share of national production. Data on firm distribution have been collected from public authorities, trade associations, and sectoral literature.108 In countries where the production of wine is marginal (Belgium, Poland, and the UK), all producers are considered to fall below this threshold. In Austria and Italy, the vast majority of producers has an output lower than 1 000 hl; however, in Austria, where production is very much atomised and there are very few large winemakers, small producers represent about 57% of the national production; in Italy, where large producers do exist, small producers only represent about 15% of national production. The only country where the number of small producers is lower than 90% is France, with 69% of wine producers estimated to be small; in terms of production, their share is in line with that of Italy (17%).

Table 12 – Estimated market share of small still wine producers and their output

AT BE FR IT PL UK EU % of wine producers 97% 100% 69% 92% 100% 100% 85% below 1,000 hl % of production 57% 100% 17% 15% 100% 100% 17% Total wine production 2,300 10 47,900 51,500 4.5 40 165,600 2015-16 (‘000 hl) Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages, from DG AGRI, AGRIMER; interviews with national trade associations, CEVI, ministries of agriculture, tax and customs authorities. Note: EU estimates based on the six sample MS.

THE FORTIFIED WINE INDUSTRY

Intermediate products are a residual category – hence, ‘intermediate’ between fermented beverages and ethyl alcohol, which includes products ‘typically based on a naturally fermented beverage to which alcohol and, in some cases, other ingredients have been added’.109 Its residual character is confirmed by the fact that it represents the least significant category in terms of tax

108 For France, data refer to wine producers below EUR 10 mn of annual revenues. A company of this size would produce less than 1,000 hl, with the exception of players producing the cheapest category of wine, i.e. without any geographical indication. Even in this case, considering 2015 prices, a company would produce about 1,250 hl. Hence, this revenue threshold is considered a good approximation of the 1,000 hl output threshold. For the UK, the share of producers below 1,000 hl is estimated based on qualitative information. 109 Commission of the European Communities, Proposal on the harmonisation of the structures of excise duties on alcoholic beverages and on the alcohol contained in other products, COM(90)432, 7.11.1990, at p.8.

140

revenues: at EU level,110 revenues from intermediate products amount to 2.3% of total excise revenues, with the maximum share in Portugal – home of Port wines – where it reaches 6.3%.

This category includes several products, such as fortified wines, vin doux naturel, vermouth, aromatised wine aperitifs, as well as certain mixed drinks with a fermented base. As a homogeneous analysis of such a diversified range of products is not possible, the focus will be on fortified wines, the most representative product in this category.111

Fortified wines are produced by adding alcohol – usually by means of a neutral strong spirit – during fermentation to increase the alcoholic strength of the product. As a result, fermentation is stopped, so that a part of the sugar content of the must is not converted into alcohol, and the resulting product has a sweeter taste.112 Port, Sherry, Madeira, Marsala, Samos and Pineau des Charentes are among the most common types of fortified wines. All these products have geographically protected indications. The quantity of fortified wines is limited when confronted to wine. As an example, the production of Port amounts to about 650 000 hl,113 the volume of Sherry is about 900 000 hl,114 in Italy about 25 000 hl of Marsala and 30 000 hl of other fortified wines are produced,115 and finally about 33 000 hl of Madeira are produced each year.116

The value chain for fortified wines include growers, producers of the base wine, and ‘fortifiers’. Grapes and base wine can be produced by a large number of wine growers: for example, 30,000 for Port,117 and 1 050 for Madeira. However, the vast majority of them does not produce the end product: fortification and ageing (where necessary) are done by shippers (also Bodegas for Sherry). For instance, 90% of the Port trade is concentrated on 15 shipping houses,118 while there are 7 producers of Madeira,119 and about 60 shippers / stockists / Bodegas for Sherry.120 The estimated average production for these producers is 39 000 hl for Port, 47 000 hl for Madeira, and 15 000 hl for Sherry. For this reason, although many small wine growers or wine makers work within the value chain, the number of small producers and their market share is residual.

110 Excluding 5 MS for which disaggregated revenues from intermediate products are not available: EL, HR, IT, MT, and PL. 111 The share of revenues from fortified wines over the total revenues from intermediate products at EU level is of 61% (sales data retrieved from IWSR, excise duty rates and revenues from EDT). The analysis excludes MT, EL, IT, PL, HR, IE and UK, because revenue data on intermediate products are not homogeneous. Outliers (NL, ES) adjusted based on EU average. 112 Court of Master Sommeliers, ‘Port – Port Trade’, Available at: www.courtofmastersommeliers.org/pdfresources/portnotes.pdf, last accessed on July 2017. 113 Correia L., Rebelo J., Caldas J., ‘Production and Trade of Port Wine: Temporal Dynamics and Pricing’, Page 16 (2012 data), 2015. 114 Great Wines from Spain, ‘The Wines – Sherry’, Available at: http://www.greatwinesfromspain.com/the-wines/sherry, 2014 Data, last accessed on July 2017. 115 Corriere Vinicolo, 2016 data. 116 Wijnstudio, ‘Madeira Wine’, Available at: http://www.madeirawine.nl/madeira-wine/, last accessed on July 2017. 117 Brito C, ‘A network perspective of the port wine sector’, International Journal of Wine, Vol. 18 No. 2, 2006. 118 The 15 members (shippers) of AEVP represent 90% of the total Port trade, Available at: http://www.aevp.pt/Members, last accessed on July 2017. 119 Discovering Madeira, ‘Who produces Madeira Wine’, Available at: http://www.discoveringmadeira.com/who-produces- madeira-wine, last accessed on July 2017. 120 Consejo Regulador de los Vinos de Jerez y Manzanilla, ‘Bodega Types’, Available at: http://www.sherry.wine/wines/bodegas, last accessed on July 2017.

141

ANNEX 11. DRIVER OF THE DYSFUNCTIONAL APPLICATION OF REDUCED RATES

Driver: Obsolete and unclear legislation

All problems relating to the application of reduced rates have their source in imprecise and obsolete provisions. The Directive was adopted 25 years ago. While the Directive works well for the main categories of alcoholic beverages, it cannot and does not cover developments which were not envisaged in 1992. These developments include:

 Increase in the number of small brewers and cross-border trade  Increase in complex business structures such as cooperative agreements  Expansion of niche products to mainstream markets (such as cider)  Increase in interest for low strength alcohol

In terms of competitive disadvantage for certain product categories, the provision of the Directive regulating which products and product categories could be subject to reduced rates, has become obsolete with time. This is due to the adoption of other EU legislation (described above), which sets the alcohol content above thresholds for certain products.

In terms of the legal uncertainty, as described under the problem definition, the Directive does not clarify what a 'legally and economically independent' brewer or distillery is. A small producer is in principle allowed to outsource, possibly under license, the production of beer to another brewer ('contract brewing'). This business relation is less common than that of a small producer brewing under license, but it may arise, e.g. when the small brewer has exhausted its production capacity. For the purpose of excise duties and rate reduction, that situation poses a number of issues. Firstly, contract brewing could be used to circumvent the output threshold (either because of the associated legal uncertainty or the lack of appropriate verification by customs authorities). Secondly, it is unclear whether contracted beer can be taxed at a reduced rate or should be treated like beer brewed under license (and thus excluded from the reduced rate). Thirdly, this contract could be seen as breaching the independence of each counterpart. If that is not the case, uncertainty remains whether each of the two economic operators individually or the jointly should remain below the output threshold in order to continue benefiting from reduced rates.121

121 A legal case is undergoing before a French court (interviews with economic operators).

142

ANNEX 12. DRIVER OF THE UNCLEAR PROVISIONS TO MEASURE THE PLATO DEGREE OF SWEETENED / FLAVOURED BEER

Driver: divergent interpretations of the term 'finished product'

The Directive does neither clarify what is a ‘finished product’ in the case of a sweetened/flavoured beer nor provide guidance on the correct method to measure its Plato degree. As a consequence, three different interpretations and measurement methods exist to determine the Plato degree of sweetened/flavoured beer (for detailed methodology see Annex 13 below). The first takes into account only the ingredients of the base beer, whereas the second and third approaches consider also the ingredients added later in the process. MS may use any of the three approaches to measure the Plato degree.

 Approach A: measures the Plato degree of the base beer, prior to the addition of sugar/flavours;  Approach B1: measures the Plato degree of the finished product after the addition of sugar/flavours taking into account only the ‘non-fermented (real) extract’, i.e. the extract of the base beer without considering sugar/flavours added to the sweetened/flavoured beer after fermentation;  Approach B2: measures the Plato degree of the finished product after the addition of sugar/flavours taking into account the ‘present extract’, i.e. the extract of the sweetened/flavoured beer including also the sugar/flavours added.

Plato is a method traditionally applied in central Europe and 14 MS calculate the excise duty using Plato degree per hectolitre of the finished product. While no complete dataset is available for the entire EU, only the sample of the MS studied in the context of the revision of the Directive show that indeed the notion of a 'finished product', and therefore the choice of the measurement method vary. Of the 50 responses to the OPC, 36 respondents interpret the term 'finished products' with reference to the 'present extract' (approach B2), which the remaining 14 interpret the term with reference to the 'real extract' (i.e. the base beer before the addition of sugars/flavours, approach A or B1). Within the beer industry, 69% of respondents agree with approach B2 and 31% interpret the term in line with approach A or B1.

143

ANNEX 13. METHODOLOGY FOR MEASURING THE PLATO DEGREE

To understand the options of measuring the Plato degree of sweetened/flavoured beer, it is useful to consider the production process, which consists of three main steps:

(1) First, the wort is created from the blending of crushed malted barley and hot water. For instance, as shown in Figure 5, 100kg of unfermented wort contain 88kg of water, 8.85kg of ‘fermentable extract’, and 3.15kg of ‘non-fermentable (or real) extract’;

(2) Second, fermentation converts part of the ‘fermentable extract’ into alcohol. The wort turns into the ‘base beer’. However, not all ‘fermentable extract’ is transformed into alcohol. In 1843, the Bohemian scientist Karl Balling found that 2.066g of ‘fermentable extract’ created 1.000g of alcohol and 1.066g of by-products (e.g. brewers’ yeast and CO2). This ratio is fixed and is hence used to calculate the Plato degree of beer. The by- product is usually removed from the base beer, which in the example below after fermentation only weighs 95.44kg.

(3) Finally, to obtain sweetened/flavoured beer, additional unfermented sugar/flavour is added to the ‘base beer’ (e.g. 3kg in the example below, which reflects the brewing process for sweetened/flavoured beer other than radler). Consequently, the sweetened/flavoured beer is heavier than the base beer (98.44kg in Figure 5). ‘Present extract’ is calculated as the sum of the ‘real extract’ and the added sugar/flavour in the sweetened/flavoured beer (6.15kg in the example).

Figure 5: The production process for sweetened/flavoured beer and determination of extract

Source: Brewers of Europe.122 Note: This example describes the production process of sweetened/flavoured beer other to which sugar/flavour is added after fermentation. Flavoured beers with additives included in the wort usually do not undergo the subsequent sweetening process. Additives included in the wort are either transformed into alcohol or part of non-fermented (real) extract. Also, the example

122 Brewers of Europe (18 May 2016), ‘European approach to calculation of Plato’, presentation

144

corresponds to the production of sweetened/flavoured beer other than radler as a relatively small amount of sugar/flavour is added. For a typical radler, the base beer is mixed with a larger quantity of lemonade (typically in proportions of 50:50), so both water and sugar are added to prepare the final product.

1. Approaches to measuring the Plato degree of sweetened/flavoured beer

Article 3(1) of the Directive requires MS to calculate the Plato degree of ‘finished products’, yet the Directive does neither clarify what is a ‘finished product’ in the case of a sweetened/flavoured beer nor provide guidance on the correct method to measure its Plato degree. So, three different interpretations and measurement methods exist to determine the Plato degree of sweetened/flavoured beer. The first takes into account only the ingredients of the base beer, whereas the second and third approach considers also the ingredients added later in the process.

1.1. Approach A: measuring the Plato degree before adding sugar/flavours

This approach aims to calculate the Plato degree of the base beer, prior to the addition of sugar/flavours. This is similar to calculating the Plato degree of non-sweetened or non-flavoured beer. In this case, one calculates the Plato degree based on the Balling formula using the real extract and mass of the base beer. The alcohol strength of the base beer in Plato degree is measured as follows:

(2.066푥퐴푙푐표ℎ표푙) + 푹풆풂풍 풆풙풕풓풂풄풕 푃푙푎푡표 푑푒푔푟푒푒 = ∗ 100 푀푎푠푠 표푓 푏푒푒푟 + (1.066푥퐴푙푐표ℎ표푙)

Therefore, following the example provided in Figure 5, the base beer is brewed at 12° Plato:

(2.066 ∗ 4.29푘푔) + 3.15푘푔 ∗ 100 = ퟏퟐ ° 푷풍풂풕풐 95.44푘푔 + (1.066 ∗ 4.29푘푔)

This approach, which is reportedly applied by Romanian authorities, focuses entirely on the features of the base beer. In fact, the quantity of water/sugar added to obtain the sweetened/flavoured beer has no impact on the Plato degree of the base beer. For tax purposes, approach A requires to apply the excise duty only to the quantity of base beer contained in the sweetened/flavoured beer. For instance, a consumer of radler including 50% of beer at 12° Plato and 50% of lemonade, would pay excise duty only on 50% of the content of the purchased bottle/can.

There is no difference between approach A and the two other approaches (B1, B2) described below with regard to sweetened/flavoured beer to which additives are included already in the wort; in such a case, the base beer corresponds to the bottled product and the real extract corresponds to the present extract.

1.2. Approach B1: measuring the Plato degree after adding sugar/flavours, on the real extract

This approach aims to calculate the Plato degree of the sweetened/flavoured product after the addition of sugar/flavours, by taking into account the ‘non-fermented (real) extract’, i.e. the extract of the base beer without considering sugar/flavours added to the sweetened/flavoured beer after fermentation, and the total mass of the sweetened/flavoured beer. The approach best reflects the actual alcohol content of the product, and is calculated as follows:

145

(2.066푥퐴푙푐표ℎ표푙) + 푹풆풂풍 풆풙풕풓풂풄풕 푃푙푎푡표 푑푒푔푟푒푒 = ∗ 100 푀푎푠푠 표푓 푏푒푒푟 + (1.066푥퐴푙푐표ℎ표푙)

In the above example, this approach yields 11.7° Plato for the sweetened/flavoured beer.

(2.066 ∗ 4.29푘푔) + 3.15푘푔 ∗ 100 = ퟏퟏ. ퟕ° 푷풍풂풕풐 98.44푘푔 + (1.066 ∗ 4.29푘푔)

1.3. Approach B2: measuring the Plato degree after adding sugar/flavours, on the present extract

This approach aims to calculate the Plato degree of the sweetened/flavoured product after the addition of sugar/flavours, by taking into account the ‘present extract’, i.e. the extract of the sweetened/flavoured beer also considering the sugar/flavours included in the sweetened/flavoured product, and the total mass of the sweetened/flavoured beer. It is calculated by applying this formula:

(2.066xAlcohol) + 퐏퐫퐞퐬퐞퐧퐭 퐞퐱퐭퐫퐚퐜퐭 ∗ 100 = 푃푙푎푡표 Mass of beer + (1.066xAlcohol)

In the above example, this approach leads to 14.57° Plato:

(2.066x4.29kg) + 6.15kg ∗ 100 = ퟏퟒ. ퟓퟕ 풅풆품풓풆풆풔 푷풍풂풕풐 98.44kg + (1.066x4.29kg)

Reportedly, this is the most used approach by tax authorities in Plato countries. Nonetheless, it is acknowledged that this approach may overestimate the Plato degree of the sweetened/flavoured beer; for this reason, the beer industry claims this method is technically incorrect. Reportedly, there is virtually no difference between method B1 and method B2 in case of artificial sweeteners (e.g. aspartame), as such sweeteners can be identified by customs lab and excluded from the calculation of the present extract123.

123 Tax authorities interviewed for the Study argued that few brewers actually use sweeteners instead of sugar, which shows the extra excise duty is not a high burden for them. By contrast, brewers explained that the choice to use sugar rather than artificial sweetener is driven by marketing considerations, e.g. using only natural ingredients, rather than by cost considerations, e.g. tax savings.

146

ANNEX 14. COUNTRY-SPECIFIC ANALYSIS OF IMPACTS OF POLICY OPTIONS FOR MEASURING PLATO DEGREE IN SWEETENED/FLAVOURED BEER

AUSTRIA

In the case of Austria the price of sweetened/flavoured beer would decrease by about 6% and consumption (volume) would increase by about 3% when switching from approach B2 (baseline and no change scenario) to approach A or B1. Tax revenues (excise duty and VAT on excise duty) generated by sweetened/flavoured beer would decrease considerably by about 43%. The changes are significant, as the Austrian sweetened/flavoured beer market consists entirely of radler. Nonetheless, when compared with total beer consumption and total tax revenue (excise duty and VAT on excise duty) on beer, the magnitude of changes becomes minor: consumption of beer would increase by only 0.2%, tax revenues would decrease by about 2%, i.e. less than EUR 5 million out of more than EUR 226 million.

Table 13: Expected impacts of the proposed policy options on the measurement of Plato degree for sweetened/favoured beer in Austria

1.B.2 / 2.B.2 (no Option 1.A / 2.A 1.B.1 / 2.B.1 change) Approach A B1 B2 (Baseline) Average price of radler 172 172 183 (EUR/hl)* % Change in price -5.9% -5.8% no change Overall consumption of 492,260 491,826 476,952 sweetened/flavoured beer (hl) % Change over 3.2% 3.1% no change sweetened/flavoured beer % Change over total beer 0.2% 0.2% no change Tax revenues from sweetened/flavoured beer 6,475,019 6,620,677 11,451,017 (EUR)** % Change over revenues from -43.5% -42.2% no change sweetened/flavoured beer % Change over revenues from -2.2% -2.1% no change total beer Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages, Author’s elaboration of IWSR and EDT series. Note: Baseline year: 2015; *Based on IWSR the Austrian sweetened/flavoured beer market consists entirely of radler; ** Excise duty and VAT on excise duty.

BELGIUM

In Belgium, the changes are weaker, given that the market is dominated by sweetened/flavoured beer other than radler. Still, tax revenues (including VAT on excise duty) fall by roughly 22- 23% when changing from approach B2 (baseline and no change scenario) to approach A or B1. Price would only decrease by about 4% for radler and 1.5% for other sweetened/flavoured beer; overall consumption of sweetened/flavoured beer would increase by some 1%. Again, such impacts appear to be marginal when compared to the overall beer consumption (+0.1%) and total tax revenues (including VAT on excise duty) on beer (-1.5%, i.e. EUR 3.5 million out of EUR 235 million).

147

Table 14 – Expected impacts of the proposed policy options on the measurement of Plato degree for sweetened/favoured beer in Belgium

1.B.2 / 2.B.2 Option 1.A / 2.A 1.B.1 / 2.B.1 (no change) Approach A B1 B2 (Baseline) Average price of radler 256 257 266 (EUR/hl) % Change in price -3.8% -3.7% no change Average price of other sweetened/flavoured beer 406 405 411 (EUR/hl) % Change in price -1.4% -1.6% no change Overall consumption of 530,552 530,885 524,948 sweetened/flavoured beer (hl) % Change over 1.1% 1.1% no change sweetened/flavoured beer % Change over total beer 0.1% 0.1% no change Tax revenues from sweetened/flavoured beer 12,341,679 12.075,829 15,795,902 (EUR)* % Change over revenues from -21.9% -23.6% no change sweetened/flavoured beer % Change over revenues from -1.5% -1.6% no change total beer Source:Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s elaboration of IWSR and EDT series. Note: Baseline year: 2015; * Excise duty and VAT on excise duty.

GERMANY

Germany has a low excise duty rate on beer (EUR 0.79/hl/° Plato in 2017). The price changes from approach B2 (baseline and no change scenario) to A/B1 are thus rather low (-2.2% for radler and about -1% for other sweetened/flavoured beer), and so are the resulting changes in overall consumption volume of sweetened/flavoured beer. Tax revenues (including VAT on excise duty) generated by sweetened/flavoured beer would change by roughly one third compared to the baseline approach. Still, the loss in tax revenue (about EUR 7 million) does not even amount to 1% of the total tax revenue from consumption of beer in Germany (more than EUR 805 million).

Table 15 – Expected impacts of the proposed policy options on the measurement of Plato degree for sweetened/favoured beer in Germany

1.B.2 / 2.B.2 (no Option 1.A / 2.A 1.B.1 / 2.B.1 change) Approach A B1 B2 (Baseline) Average price of radler 184 184 189 (EUR/hl) % Change in price -2.2% -2.2% no change Average price of other sweetened/flavoured beer 229 229 231 (EUR/hl) % Change in price -1.0% -1.2% no change

148

Overall consumption of 1,914,662 1,914,542 1,894,811 sweetened/flavoured beer (hl) % Change over 1.0% 1.0% no change sweetened/flavoured beer % Change over total beer 0.0% 0.0% no change Tax revenues from sweetened/flavoured beer 12,779,642 12,794,961 19,806,253 (EUR)* % Change over revenues from -35.5% -35.4% no change sweetened/flavoured beer % Change over revenues from -0.9% -0.9% no change total beer Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s elaboration of IWSR and EDT series. Note: Baseline year: 2015; * Excise duty and VAT on excise duty.

ITALY

Ad discussed above, Italy is undergoing a regulatory review process that embraces also the method for the measurement of Plato degree for excise duty purposes. According to some stakeholders, this may eventually result in a transition from approach B2 to B1, but since the competent authority has not yet adopted the secondary implementing regulation the outcome of the process is still uncertain (very likely a decision will be taken after the issuance of the CJEU judgement on the Polish case). Whereas taxes are currently computed based on approach B2 in this simulation Economisti Associati assumed that the country has completed its transition to approach B1. It is important to highlight that this is a hypothetical assumption made for analytical purposes. Under this assumption, approach B1 is the baseline (and ‘no change’ scenario) while approach A and B2 are the ‘change scenarios’.

Due to the country’s relatively high excise duty rate (3.04 per hl/Plato degree) and the narrow market for sweetened/flavoured beer, which is dominated by radler, changing the measurement approach results in rather high percentage variations in tax revenues (excise duty and VAT on excise duty) generated by sweetened/flavoured beer when moving back from approach B1 to B2 (+72%). Changes from approach B1 to A are rather minor.124 Interestingly, in light of the very limited size of the Italian market for sweetened/flavoured beer, any change in consumption and tax revenues is marginal compared to the entire market for beer.

Table 16 – Expected impacts of the proposed policy options on the measurement of Plato degree for sweetened/favoured beer in Italy

1.B.1 / 2.B.1 (no Option 1.A / 2.A 1.B.2 / 2.B.2 change) B1 (hypothetical B2 (Current Approach A dynamic baseline situation) scenario)

124 As discussed previously, as the Italian authorities are still in the process of setting secondary rules to complete the transition from approach B2, it is still also possible a transition from B2 to A rather than to B1. In this respect, the impact analysis in Table 16 confirms that approaches A and B1 lead to very similar results in Italy. Therefore, the findings of the impact analysis performed in the Study remain largely valid, irrespective of whether Italy will eventually opt for approach A or B1.

149

Average price of radler 238 238 254 (EUR/hl)* % Change in price -0.2% no change 6.8% Overall consumption of 217,481 217,246 209,213 sweetened/flavoured beer (hl) % Change over 0.1% no change -3.7% sweetened/flavoured beer % Change over total beer 0.0% no change 0.0% Tax revenues from sweetened/flavoured beer 4,420,685 4,519,242 7,762,097 (EUR)** % Change over revenues from -2.2% no change 71.8% sweetened/flavoured beer % Change over revenues from 0.0% no change 0.4% total beer Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s elaboration of IWSR and EDT series. Note: Baseline year: 2015; *Based on IWSR the Italian sweetened/flavoured beer market consists entirely of radler;125 ** Excise duty and VAT on excise duty.

POLAND

In Poland, where tax revenues are the highest among sampled countries due to the larger size of the market for sweetened/flavoured beer, the changes are modest in absolute terms, as the market is dominated by flavoured beer other than radler and the national excise duty rate is moderate (EUR 1.86/hl/° Plato in 2015). Tax revenues (excise duty and VAT on excise duty) generated by sweetened/flavoured beer would fall by more than EUR 15 million when changing from approach B2 (baseline and no change scenario) to A or B1, i.e. about -1.5% when compared to total tax revenues on beer (more than one billion EUR). Impacts on consumption are more limited (-2.6% over consumption of flavoured/sweetened beer; -0.2% over total consumption of beer).

Table 17 – Expected impacts of the proposed policy options on the measurement of Plato degree for sweetened/favoured beer in Poland

1.B.2 / 2.B.2 (no Option 1.A / 2.A 1.B.1 / 2.B.1 change) Approach A B1 B2 (Baseline) Average price of radler 99 100 110 (EUR/hl) % Change in price -9.4% -9.2% no change Average price of other sweetened/flavoured beer 198 197 203 (EUR/hl) % Change in price -2.9% -3.3% no change Overall consumption of 2,445,951 2,448,589 2,384,762

125 In Italy there is a small, but declining market for flavoured beer, which appears not to be recorded by IWSR data. Nonetheless, the Italian market for flavoured beer other than radler is dominated by beer with addition of flavour in the wort produced by craft brewers; the Plato degree of such beer is not affected by different measurement approaches. Hence, IWSR data allows capturing the entire market relevant to the policy problem.

150

1.B.2 / 2.B.2 (no Option 1.A / 2.A 1.B.1 / 2.B.1 change) Approach A B1 B2 (Baseline) sweetened/flavoured beer (hl) % Change over 2.6% 2.6% no change sweetened/flavoured beer % Change over total beer 0.2% 0.2% no change Tax revenues from sweetened/flavoured beer 56,537,824 55,445,881 72,404,261 (EUR)* % Change over revenues from -21.9% -23.4% no change sweetened/flavoured beer % Change over revenues from -1.5% -1.6% no change total beer Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s elaboration of IWSR and EDT series. Note: Baseline year: 2015; * Excise duty and VAT on excise duty.

ROMANIA

Finally, Romania would see no change if approach A were selected. Switching to from approach A (baseline and no change scenario) to approach B1 would make almost no difference, whereas switching to approach B2 would result in a 2.3% decrease in consumption of sweetened/flavoured beer (price of radler would increase by 5%; price of other sweetened/flavoured beer by 1%) and a 56% increase in tax revenues (including VAT on excise duty) generated by this type of beer. However, this corresponds to only -0.1% in total beer consumption and +0.8% in total tax revenue from excise duty on beer in Romania (i.e. less than EUR 2 million out of more than EUR 195 million).

Table 18 – Expected impacts of the proposed policy options on the measurement of Plato degree for sweetened/favoured beer in Romania

1.A / 2.A Option 1.B.1 / 2.B.1 1.B.2 / 2.B.2 (no change) Approach A (Baseline) B1 B2 Average price of radler 92 93 97 (EUR/hl) % Change in price no change 0.1% 5.1% Average price of other sweetened/flavoured beer 222 221 225 (EUR/hl) % Change in price no change -0.2% 1.2% Overall consumption of 369,663 369,509 361,127 sweetened/flavoured beer (hl) % Change over no change 0.0% -2.3% sweetened/flavoured beer % Change over total beer no change 0.0% -0.1% Tax revenues from sweetened/flavoured beer 2,692,520 2,700,865 4,190,476 (EUR)* % Change over revenues from no change 0.3% 55.6% sweetened/flavoured beer % Change over revenues from no change 0.0% 0.8%

151

total beer Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s elaboration of IWSR and EDT series. Note: Baseline year: 2015; * Excise duty and VAT on excise duty.

152

ANNEX 15. ASSESSMENT OF THE IMPACT ON THE MARKET OF THE RE-CLASSIFICATION OF CERTAIN PRODUCTS

The re-classification of certain products under a different tax category with a different excise duty rate would clearly have an impact on the market size and trends. This impact has been assessed triangulating the results of a quantitative market analysis126 econometric model with other evidence collected through interviews with stakeholders and a desk review of literature and relevant documentary sources. The various steps of the assessment and the findings are described in the following paragraphs and summarised in Table 19, Table 20, and Table 21 below.

Step 1 – estimating the market size of potentially affected products. This part of the exercise was conducted as part of the baseline assessment and led to the quantification of the market size of both ‘borderline’ products (mixed drinks and other fermented beverages that might have lost their essential fermented character) and other sub-categories of products that might be unintendedly affected since currently covered by the same Directive provisions at stake (Article 12 or Article 17). This part of the work required a detailed one-by-one analysis of a vast range of specific brand-products listed in the IWSR database. The tax treatment of these products is not disclosed so it had to be inferred from e.g. alcoholic strength, estimated alcoholic base and, in some cases, market price.

The sales volume of products with similar characteristics and falling in the same (estimated) tax category were then aggregated into homogeneous sub-categories considered relevant for the assessment. The target products include non-spirit mixed-drinks with ABV lower than 5.5% vol or lower than 10% vol, and other medium/high strength fermented beverages with an ABV up to 22% vol. Non-target products consist of certain aromatised-wine products that are possibly taxed under Article 12, including both CN 2206 and certain CN 2205 products. In practice, non- target products include product like sangria, gluehwein (mulled wine), and other aromatised- wine cocktails. Economisti Associati also estimated the market of ‘borderline’ cider, in case it would be included in the scope of the re-classification, in line with the approach outlined in Section 2.1. Overall, it is estimated that ‘target products’ sales in the EU amount to approximately 154 mn litres (305 mn litres if ‘borderline’ cider is included127); and non-target products potentially affected to some 106 mn litres. As compared to the total volume of alcoholic beverages consumed in the EU per year, the products at stake are only a tiny minority, i.e. 0.8% of the total.

The sub-categories are defined assuming that all the encompassed products would change of tax category following the adoption of one of the policy options considered. Since the attribution of products to a certain sub-category, as well as the very market dimension of these products are subject to a certain margin of error, the baseline data used in the analysis should be taken with caution.

126 The market analysis is largely based on the results of an econometric modelling exercise conducted on the market data published by IWSR. 127 Where not explicitly mentioned, the figures ‘borderline’ cider does not include the UK market. The rationale is that since the UK market accounts alone to two-thirds of the EU cider, and is also much greater than the total market of mixed drinks, it may ultimately determine the results of the for impact assessment exercise. Secondly, in the current situation it is very unlikely that the UK would follow the EU in a re-classification process that would affect primarily the competitiveness of its domestic cider industry.

153

Step 2 - market trends in the absence of policy changes. This step consisted in estimating the projected value of the market after one year in the ‘business as usual’ scenario (i.e. dynamic baseline). Projections are based on the average growth rate observed over the past five years. Economisti Associati limited the market projections assuming that growth trend is linear, and would remain similar also in the following years. The trend varies across sub-categories of product: ‘borderline’ OFB and IP, as well as AWP - CN 2206 are substantially stable, AWP – CN 2205 seems declining, while ‘borderline’ cider is growing.

The overall market change is positive, although very modest in scale. The aggregated annual variation for these sub-categories amount to less than 1.0 mn litres, which is some +0.2% per year. Needless to say, growth trends differ across the MS that have been examined and used to extrapolate general EU-level trends.

Step 3 - revised tax rate due to re-classification. Option I may entail in practice that a certain amount of products with a questionable essential fermented character are taxed in accordance to Article 20. These may regard both certain Article 12 products (mixed drinks) and, more likely, certain Article 17 products – i.e. medium/high strength OFB currently considered as Intermediate Products. It is unlikely that any cider including ‘borderline’ ones could be affected. Also CN 2205 products would not be affected, since this option concerns only CN 2206 products. Instead, some AWP classified as CN 2206 may in theory (but not very likely) be affected. Similar outcomes could be obtained through non-regulatory options III.a and III.b.

For Option II Economisti Associati have not envisaged any specific tax rate, since this is outside of the scope of this exercise. However, for illustrative purposes, Economisti Associati simulated that of the two categories created by splitting the current OFB category, one would retain the current Article 12 tax rate, while the other would be taxed with the same rate of IP. The simulation was conducted at the level of each of the six sample MS taking into account the different rates currently applied to OFB and IP and the existence of national non-harmonised measures (e.g. pre-mix tax, separate excise duty for cider, application of Article 17(2) etc.).

Step 4 - tax-induced variation of the current price levels. An important variable of the econometric model applied is the extent to which a possible variation in the excise duty rate may translate into a variation of the average price level for a certain sub-category. This variable – denominated ‘pass-through’ factor – has been calculated for all the categories of products potentially concerned by re-classification (intended and unintended ones) based on a large matrix of historical correlation between tax (including excise duty and the VAT on the excise duty) and price levels in the six sample MS. In practice, the ‘pass-through’ factor expresses by how much the selling price of a product would change following a variation of the tax levied. It is important to remind that the tax level is only one of the possible explanatory factors behind the price level. As most of the interviewees highlighted, prices are only limitedly influenced by taxes and more importantly by marketing strategies, production costs, retail mark-ups, etc.

Fermented mixed drinks below 10% vol are the only case where no statistically-relevant correlation could be found. More than for other beverages, the price of mixed drinks seems therefore determined by factors other than the tax level. While this is the case for ordinary and modest variations of rates, in the case of special taxes (pre-mix or alcopop taxes) explicitly conceived to deter consumption, major market impacts were indeed observed, consisting in the

154

massive withdrawal of affected products from the market. So, in the model, Economisti Associati assumed for these products a conventional pass-through of 100%.128

Step 5 - overall variation in the demand. The main outcome of the exercise consisted in estimating the variation of consumers’ demand of products possibly caused by the application of a different tax rate - taking into account the above effects on prices. This required in the first place to calculate the elasticity of the demand for the various sub-categories of products to re- classify. Economisti Associati used for this purpose the same large dataset of Step 4. Combining the estimated variation in price levels (Step 4) and the elasticity coefficient, it was eventually possible to estimate the variation in the volumes of product demanded potentially caused by the two regulatory options at stake.129 For a more accurate estimation, two different econometric models have been applied to data, which returned partly different results (but coherent in terms of general trends). As shown in Table 21, the two models produced a minimum and a maximum impact scenario. The ‘mean’ value between the two endpoints can be taken as a valid approximation.

The assessment of impact has been conducted on the six sample MS, and the outcomes were extrapolated at EU level by applying appropriate conversion factors linked to market size. These are smaller in the case of mixed drinks, IP and AWP – where this sample accounts for some 47% of the EU market, and bigger for cider – where they represent only 14%. The principle behind extrapolation is that the sample is sufficiently representative of the entire EU market, not only in quantitative terms but also qualitatively, and in particular that the variety of preferences and trends observed in the sample sufficiently reflects the diversity of EU countries. There are a couple of limitations in this method that are worth mentioning: (i) the outcomes of the exercise aim at representing the expected EU-aggregated impacts, but do not support conclusions on impact on individual MS; (ii) the ratio between this sample and EU-level data changes if instead of volume of products (in litres) Economisti Associati consider the value of market (in EUR) or the amount of excise duty collected. The extrapolation of results for these other variables using the volume of consumption as conversion factor inevitably leads to minor calculation distortions that could not be entirely corrected.

The results presented in Table 19 below show that Option I would affect primarily ‘borderline’ IP, with a possible reduced volume of sales of ca. -36% (average value between ‘min’ and ‘max’ scenarios).130 Mixed drinks of lower strength would also be affected, but at a smaller degree, since the structure of Article 20 is by pure alcoholic degree. Overall, the consumption of target products would reduce by some 42 mn litres in one year (average scenario). The collapse is mostly due to the abrupt introduction of a relatively high excise duty on products that in various markets currently enjoy a zero or very low excise duty. It is also due to the fact that the demand

128 Applying a different ‘pass through’ factor to mixed drinks, the model would evidently return different estimates. In the interim stage of the work, we had calculated the impact applying a greater pass-through coefficient (1.5 - prices increase in a greater proportion than the tax increase) and a smaller one (0.5). In the first case, the impact on mixed drinks were magnified, while in the second case they were mitigated. The two alternative coefficient used were arbitrary, so the results had little analytical significance. In the final version of the Study, we have approached the issue of sensitivity of results by using two different econometric models, and calculating an upper and a lower threshold to the estimates provided. Therefore, the less-sophisticated simulations by different values of the pass-through coefficient have been dropped. 129 The ‘arc elasticity’ formula has been used in this exercise, in consideration of the fact a big variation is expected on a category of products with varying starting prices and sales quantity, and given the absence of a specific demand function for these products. In practice, as compared to basic ‘point elasticity’, the arc elasticity defines the mid-point elasticity between the two selected points and may mitigate somehow the overall effects. 130 Where not specified all figures in this section refer to the average value between the minimum and maximum scenarios provided in Table 19.

155

of these products is quite elastic, so the consumers would likely respond to a price increase turning massively to other products.

The impact of Option II would be borne in particular by very low-strength mixed drinks and – if included in the re-classification – by ‘borderline’ cider. The model predicts a sales drop of respectively 46% (for very low-strength mixed drink – average scenario) and 64% (for ‘borderline’ cider – average scenario). More moderate is the expected impact on mixed drink between 5.5% and 10% vol, which in some MS are already taxed as Intermediate Products. The aggregated market loss would be greater than under Option I, i.e. – ca. 91 mn litres, primarily due to the ‘flat’ nature of the excise duty that would applied, whose burden is inversely proportional to the ABV strength.

It is important to highlight the estimated effects on non-target products. Under Option I some aromatised wine products classified as CN 2206 may unintendedly fall in the scope of reclassification. In this case, applying their corresponding pass-through factors and elasticity coefficient, Economisti Associati expect a reduction of sales from ca. 36 mn litres to nearly zero. The variation would be much greater than for target products. Under Option II, the impact on non-target products would be equally profound. Adverse market effects may be registered also by some CN 2205 products currently in the remit of Article 12. Overall, the AWP segment may register a drop of -74 mn litres (average scenario), i.e. some -70% against the ‘no change’ scenario.

As discussed, a quantitative assessment of the impact of non-regulatory options (in particular options III.a and III.b) would be highly speculative, since these options fall outside of the remit of excise duty system and/or are non-binding in nature. Nonetheless, since these options would essentially clarify the conditions under which certain fermented beverages should be treated like spirits, it can be assumed that their impact is conceptually similar to regulatory Option I. This is even more so, since Option I de facto requires that operational guidelines are adopted in support to the regulatory amendment.

Table 19 – Estimated impacts of the proposed options on market sales volume Product Baseline Baseline Option categories 2016 + 1 year Range Option I Diff. II Diff. (mn (mn (mn (mn (mn litres) (mn litres) (*) litres) litres) litres) litres) ‘Borderline’ OFB Mixed Drinks max 54.62 -19.73 27.92 -46.42 73.64 74.34 (<= 5,5% vol) min 65.87 -8.47 51.81 -22.53 Mixed Drinks max 4.45 -0.59 4.54 -0.50 4.76 5.04 (5,5%-10% vol) min 4.79 -0.25 4.82 -0.22 ‘Borderline’ cider (OFB) max 459.44 0.00 97.14 -362.30 w/ the UK 435.27 459.44 min 459.44 0.00 221.52 -237.91 max 153.62 0.00 33.80 -119.82 w/out the UK 151.80 153.62 min 153.62 0.00 77.08 -76.54 ‘Borderline’ IP MHS Ferm. max 40.99 -34.50 75.48 0.00 75.49 75.48 (10% - 22% vol) min 54.50 -20.98 75.48 0.00 Non-target products** max 68.15 0.00 29.28 -38.87 AWP 2205 70.12 68.15 min 68.15 0.00 24.73 -43.42 max 0.50 -35.93 0,00 -36.43 AWP 2206 36.43 36.43 min 0.61 -35.82 5.59 -30.85 max 322.32 -90.75 171.03 -242.04 TOTAL 412.23 413.07 min 347.54 -65.53 239.52 -173.56

156

mean 334.93 -78.14 205.27 -207.80 (w/ UK cider)*** (695.70) (718.88) (640.75) (-78.14) (309.16) (-409.72)

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author’s estimates, based on a quantitative analysis of IWSR data. Notes: The ‘baseline + 1 year’ is estimated on the basis of 2016 data and the average growth rate registered in the past five years. (*) For each data point a maximum and a minimum impact is provided, based on the results of the two econometric models used in the Study. (**) Non-target products that might be affected by the policy options include aromatised-wine products (AWP) falling under CN 2206 or, in some circumstances, under CN 2205. (***) Since the UK cider market account for the bulk of EU cider, the impact on this market might determine alone the overall results of the exercise. For this reason, the aggregated figures containing the UK cider are provided separately.

Step 6 - overall effects on market value. The reduction in sales has eventually been combined with the tax-induced expected increase in prices in order to estimate the scale of the impact in terms of market value (Table 20). These are evidently negative due to the expected market decline. In both ‘change scenarios’ considered, the estimate loss would be around EUR 300-400 mn.

These figures have to be considered in the light of an overall EU28 market that according to IWSR amounts to EUR 207.2 bn. In this respect, the products at stake (target and non-target) represent altogether a small 1%, and the possible value loss would be of 0.2%. Also, it has to be considered that the consumption would likely shift to other products, so at systemic level the variation would be hardly noticeable.

Table 20 – Estimated impacts of the proposed options on market economic value

Product Baseline Baseline Rang categories 2016 + 1 year e Option I Diff. Option II Diff. (€ mn) (€ mn) (*) (€ mn) (€ mn) (€ mn) (€ mn) ‘Borderline’ OFB Mixed Drinks max 387.11 -123.19 236.05 -274.25 513.45 510.30 (<= 5,5% vol) min 459.61 -50.69 443.94 -66.36 Mixed Drinks max 38.38 -7.76 39.02 -7.12 43.74 46.13 (5,5%-10% vol) min 41.63 -4.51 41.68 -4.46 ‘Borderline’ cider (OFB) max 2,790.87 0.00 732.70 -2,058.17 w/ the UK 2,644.07 2,790.87 min 2,790.87 0.00 1,663.81 -1,127.06 max 478.84 0.00 254.95 -223.89 w/out the UK 473.16 478.84 min 478.84 0.00 578.95 100.11 ‘Borderline’ IP MHS Ferm. max 533.81 -333.35 867.17 0.00 868.43 867.17 (10% - 22% vol) min 689.96 -177.21 867.17 0.00 Non-target products** max 200.89 0.00 107.52 -93.37 AWP 2205 210.37 200.89 min 200.89 0.00 92.24 -108.65 max 3.40 -9.05 2.85 -9.60 AWP 2206 12.45 12.46 min 4.19 -8.27 28.69 16.24 max 1,642.43 -473.36 1,507.56 -608.23 TOTAL 2,121.61 2,115.79 min 1,875.10 -240.69 2,052.67 -63.12 mean 1,758.77 -357.02 1,780.11 -335.67 (w/ UK cider)*** (4,292.52) (4,427.82) (4,070.80) (-357.02) (2,561.42) (-1,866.40)

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; author's estimates, based on a quantitative analysis of IWSR data. Notes: The ‘baseline + 1 year’ is estimated on the basis of 2016 data and the average growth rate registered in the past five years.

157

(*) For each data point a maximum and a minimum impact is provided, based on the results of the two econometric models used in the Study. (**) Non-target products that might be affected by the policy options include aromatised-wine products (AWP) falling under CN 2206 or, in some circumstances, under CN 2205. (***) Since the UK cider market account for the bulk of EU cider, the impact on this market might determine alone the overall results of the exercise. For this reason, the aggregated figures containing the UK cider are provided separately.

Table 21 – Estimated impacts of the proposed options on tax revenues (excise duty and the VAT applied to it) Diff. Diff. Diff. (w/ Diff. Product Baselin Baseline Ran (w/ (only Option VAT) (only ED) categories e 2016 + 1 year ge Option I VAT) ED) II (€ mn) (€ mn) (*) (€ mn) (€ mn) (€ mn) (€ mn) (€ mn) (€ mn) ‘Borderline’ OFB max 137.21 -76.44 -62.92 100.73 -112.91 -92.93 MD very low 214.97 213.65 min 170.86 -42.78 -35.21 213.94 0.29 0.24 max 6.48 -1.56 -1.28 5.99 -2.05 -1.69 MD low 7.62 8.04 min 7.30 -0.74 -0.61 6.79 -1.25 -1.03 ‘Borderline’ cider (OFB) max 871.10 0.00 0.00 212.06 -659.04 -542.42 w/UK 825.28 871.10 min 871.10 0.00 0.00 542.48 -328.62 -270.47 max 348.37 0.00 0.00 73.79 -274.58 -225.99 W/out UK 344.24 348.37 min 348.37 0.00 0.00 188.76 -159.61 -131.36 ‘Borderline’ IP max 137.17 -87.87 -72.32 225.03 0.00 0.00 MHS Ferm. 224.43 225.03 min 189.51 -35.52 -29.24 225.03 0.00 0.00 Non-target products** max 7.67 0.00 0.00 37.03 29.36 24.16 AWP 2205 8.15 7.67 min 7.67 0.00 0.00 31.42 23.75 19.55 max 0.32 -3.47 -2.85 0.00 -3.78 -3.11 AWP 2206 2.61 3.78 min 0.39 -3.39 -2.79 11.23 7.45 6.13 - max 637.21 -139.37 442.57 -363.97 -299.57 169.34 TOTA 802.01 806.54 min 724.10 -82.44 -67.86 677.17 -129.37 -106.48 L mea - 680.65 -103.61 559.87 -246.67 -203.02 n 125.89 (w/ UK (1,283. (1,329.2 (1,203.3 (- (- (805.87) (-523.41) (-430.79) cider)*** 05) 7) 8) 125.89) 103.61) Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages"; Author’s estimates based on a quantitative analysis of IWSR data. Notes: ED: Excise duty (revenue); w/ VAT: excise duty augmented with the applicable VAT. The average EU28 VAT rate is conventionally applied (21.5%). The ‘baseline + 1 year’ is estimated on the basis of 2016 data and the average growth rate registered in the past five years. (*) For each data point a maximum and a minimum impact is provided, based on the results of the two econometric models used in the Study. (**) Non-target products that might be affected by the policy options include aromatised-wine products (AWP) falling under CN 2206 or, in some circumstances, under CN 2205. (***) Since the UK cider market account for the bulk of EU cider, the impact on this market might determine alone the overall results of the exercise. For this reason, the aggregated figures containing the UK cider are provided separately.

158

ANNEX 16. PROBLEMS THAT WILL NOT BE ADDRESSED WITHIN THIS IMPACT ASSESSMENT

The excise duty exemption for private production of fermented beverages (i.e. beer, wine and other fermented beverages (OFB)) for home consumption, which was reviewed in the Ramboll Evaluation, will not be further considered in this impact assessment for the reasons explained below.

All stakeholders consulted and data analysed suggest that the private production of fermented beverages does not create any significant problems to the internal market, either economic or regulatory. Public authorities reported no cases where home brewing was linked to tax frauds and there are no indications of any competition distortion.

Private production of ethyl alcohol131 and intermediate products (IP) was not granted in 1992 for both health and tax reasons. Distillation is more dangerous than fermentation from a health perspective and given the higher excise duties on ethyl alcohol, the risk of tax fraud is higher. According to the Study, the amount of illicit private distillation is estimated to be low and likely to decrease in the future. This decrease is due to several factors which are:

 A decline in total alcohol consumption,  Lower number of people living in rural areas,  Changes in consumers' lifestyle and preferences,  Increase in disposable income,  An increase in the accessibility of alcohol, and  The loss of traditional production methods and techniques.

The Study considered the possibility of extending the optional exemption for private production and home consumption to ethyl alcohol and IP to address this discrimination. The extension of this optional exemption to the private production of ethyl alcohol and IP would have modest but negative impacts on tax revenues and market effects. 54% and 38% of respondents to the OPC opposes the extension of possible tax exemptions to the private production of ethyl alcohol and IP, respectively. The number of respondents in favour of the extension was, for both categories, slightly less than one third of the total, with the balance of respondents expressing a neutral position. MS where private distillation is not allowed maintain that the situation should not change for the following reasons: 1. The health risks associated with methanol poisoning, 2. The possible increase in the consumption of spirits due to the liberalisation of private distillation, 3. Ethyl alcohol presents a higher risk of frauds.

The Council mandate132 explicitly called for the right balance to be struck between an extension of the exemption to all alcoholic beverages and the risk of (unintended) negative effects. Therefore based on the feedback from the stakeholders and the Council mandate, the exemption for private production is not taken into account in this impact assessment.

131 Private distillation is possible in Austria and Romania. These provisions reportedly find their justification in the minutes of the Council meeting at which the Directive was adopted, stating that MS were allowed to maintain 'traditional exemptions' for the private production of any alcoholic beverage. 132 'Council Conclusions on the Commission Report to the Council on the evaluation of Council Directive 92/83/EEC on the structures of excise duties on alcohol and alcoholic beverages', 06.12.2016

159

ANNEX 17. COMPARISON OF IMPACTS OF ALTERNATIVE OPTIONS

Dysfunctions in the application of exemptions for denatured alcohol

Table 22 - Comparison of impacts addressing option on mutual recognition of CDA Impact area and 0 - No Change 1 - Clarify mutual recognition target groups Reduction of any remaining trade Functioning of the Most problems resolved by 0 / barriers and distortions due to possible Single Market and 0 adoption of CIR 2017/2236; no +1 restrictive interpretation of mutual competition further change. recognition by some MS. No impact on most businesses, as this would only codify the approach already Most problems resolved by taken by most MS. Operating costs and 0 / 0 adoption of CIR 2017/2236; no Minor positive impacts for producers conduct of business +1 further change. that sell CDA to MS with different national formulations, and users of CDA in these MS. Enforcement costs for national authorities 0 No change. 0 No change.

Fiscal fraud and Risk of fraud with CDA reduced associated revenue, 0 significantly by adoption of CIR 0 No change. health risk 2017/2236; no further change. Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017; Legend: +2 major positive effect expected; +1 moderate positive effect expected; 0 no effect or neutral impact expected; -1 moderate negative effect expected; -2 major negative effect expected.

Table 23 – Comparison of impacts of options addressing problems stemming from the proliferation of national approaches to PDA Impact area 2 - Partial 3 - Confidence / 4 – Legal clarification of and target 0 - No Change harmonisation of PDA capacity building provisions relating to PDA groups formulations measures Highly uncertain – More equal treatment of Reduced barriers to 0 may lead to goods containing PDA Functioning intra-EU trade, reduced barriers if + Equality of treatment of of the Single No change fairer competition 0 +1 / MS adopt more 1 PDA for indirect uses Market and expected between PDA + consistent rules / across the EU competition producers and users 1 practices as a in different MS result Benefits for PDA Cost savings for users of producers and users PDA in MS that currently that operate in more do not exempt indirect than one MS Highly uncertain – uses. 0 Operating Possible positive or may lead to Lower risk of delays / costs costs and No change negative effects for reduced costs if 0 associated with disputes 0 +1 / conduct of expected users depending on MS adopt more / with customs. + business whether the consistent rules / + Potential increases in 1 harmonised list is practices 1 movement cost for a more or less limited number of products exhaustive than the current national one

160

Impact area 2 - Partial 3 - Confidence / 4 – Legal clarification of and target 0 - No Change harmonisation of PDA capacity building provisions relating to PDA groups formulations measures EU funding via Short-medium term: the Fiscalis significant resources programme required for - Enforcement MS human and -1 developing 1 costs for No change financial resources 0 / harmonised list 0 No change expected national expected May lead to +1 Medium-long term: / authorities savings if MS cost savings for 0 adopt more authorities incl. efficient rules / laboratories practices Reduced risk of Adoption of fraud involving Fiscal fraud Eurodenatur 0 Highly uncertain – 0 products containing Reduced scope for and ant for CDA may lead to / 0 / ‘weakly’ denatured + intentional associated may displace / reduced risks if - +1 alcohol (if ‘low 1 misclassification of PDA revenue, fraud + MS adopt stricter 1 fiscal risk’ criterion so as to avoid controls health risk towards 1 rules / practices is implemented PDA strictly) Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017; Legend: +2 major positive effect expected; +1 moderate positive effect expected; 0 no effect or neutral impact expected; -1 moderate negative effect expected; -2 major negative effect expected.

161

Dysfunctions in the classification of certain alcoholic beverages

Table 24 – Comparison of impacts: review of the scope of OFB category I) Clarify the excise duty structure for II) Introducing a differentiation in the III & IV) Other approaches not Impact area 0) No Change ‘borderline’ products including AFC OFB tax category requiring a revision of the Directive containing products Re-classification would lead to a significant reduction in sales of 0 certain ‘borderline’ products Other factors than taxes influence 0 / especially in the category of Market impact depends on the tax market and competition more +1 ‘borderline’ IP, redressing some rate applied to the new category: an pervasively. CN / CNEN review and guidelines apparent malfunctioning. IP-like treatment would lead to a Very limited cases of competition may have similar benefits and However, several non-target CN collapse of low strength mixed Competition distortion reported. No change drawbacks as (I). 2206 products would be unintendedly 0 / drinks, ‘borderline’ cider as well as 0 / and market expected. affected. -1 other non-target products. +1 effects Sectoral legislation for cider may Drawing a demarcation line between It is possible (not demonstrated) that reduce the risk of competition More clarity and predictability may ‘traditional’ cider and ‘mass-market’ the ambiguity of the text constrains issues. support AFC market growth. products is sensitive and may easily market development for AFC 0 / - 0 / - In absolute terms, the volume of result in competition issues. products in certain MS. 1 1 products at stake is very modest, so limited change in the overall market would be perceived. Tax revenues have kept increasing +1 CN / CNEN review and guidelines and the magnitude of the issue The net effect on tax revenues is are likely to have the same impact potentially caused by borderline moderately negative due to the as option (I). products is modest and declining. estimated elasticity of demand. 0

No relevant change expected if the No losses are actually expected due 0 Enhanced monitoring and control treatment of AFC in products is to substitution with other products Risk of losses if the tax rate applied 0 (using EPC): clarified. that would likely occur. is high (a ‘per ABV’ structure would Tax The risk of ‘misclassifications’ and Enhanced monitoring and control -1 have more balanced impacts). revenues ensuing tax losses would be (using EPC): No major changes expected. As for option (I) substitution would 0/ bridged. In the few MS with different excise The adoption of a fixed threshold - mitigate losses. -1 Furthermore, there would be more rates for wine and OFB, the risk of higher than the strictly necessary clarity in the tax treatment as ‘misclassifications’ may translate dose - may translate in an excessive +1 Article 8 or Article 12 of certain into incorrect excise duty levied and amount of alcohol (AFC) that is not aromatised wine products, useful potential loss. The issue would be taxed as ‘ethyl alcohol’. -1 for market monitoring purposes. magnified by a possible adoption of a

separate tax category for certain OFB. CN / CNEN review and guidelines: Like option (I), but with reduced +1 initial costs, since no action at the

level of the excise duty system is

required.

Limited in absolute terms and Negative in the short term due to Negative in the short term (as option declining, thanks to the adoption of Enhanced monitoring and control: Administrati one-off initial costs. I), with extra costs envisaged for MS level approaches. -1 / Legal and technical revisions ve burden +1 Positive in the long term due to -1 updating the system. +1 required, both for economic reduction of the burden to deal with Not so effective in reducing the Enhanced monitoring and control: operators and competent complex cases. burden from complex cases. not relevant authorities.

If the required changes are limited

to the OFB economic operators

directly concerned the level of -1 administrative burden would remain modest.

Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017, based on a quantitative analysis of IWSR data.

Dysfunctional application of reduced rates

Table 25 – Comparison of impacts of reduced rates for small brewers

No change Option 1.a – normalising the Option 1.b – creating conditions for Impact area and definition of economic and legal recognition of small brewers across target groups independence at EU level the EU

Administrative The reduced rates for A small increase of burdens for The administrative burdens small brewers do not administrative burdens could be economic 0/ would be minimal generate unnecessary expected in case a uniform operators +0 -1 irrespective of the 0/ -1 administrative certificate for small brewers is approach (a non-binding burdens and no introduced. Impact is estimated instrument or a legislative evolution is expected. at 7.5% of the current burdens. revision.

Enforcement Enforcement costs are Public authorities would incur costs for public considered to be The enforcement costs additional costs if the ex-ante authorities minimal by tax and would be minimal uniform certificate is adopted. customs authorities. 0/ - irrespective of the Enforcement costs would be Increased complexity -1 -1/ -2 1 approach (a non-binding concentrated in countries not and cross-border instrument or a legislative having implemented the flows may lead to revision. reduced rates for small small incremental economic operators. costs.

SME More legal clarity and ease of competitive-ness An improvement in the doing business for cross-border More players are legal clarity would have a economic operators would likely to benefit from positive impact on SME improve the competitiveness of 0 / the reduced rates, competitiveness. However, SME, and facilitate the +1 +1 +1 given the growth of the magnitude could differ consolidation of medium the small brewery if MS decide not to players. However, given the market segment. conform to the non-binding limited scale of the problem, guidelines. positive impacts are expected to be modest.

Cross-border An improvement in the A more uniform approach to the The expected increase market effects for legal clarity would have a application of reduced rates to in cross-border flows economic positive impact on the small brewers would facilitate may lead to a modest operators cross-border functioning of the ease of doing business for 0 / increase in the +1 the scheme. However, the +1 cross-border economic -1 impacts of the minor magnitude could differ if operators. However, given the disturbances to the some MS decide not to limited scale of the problem, Single Market conform to the non-binding positive impacts are expected to identified. guidelines. be modest. Legend: +2 major positive effect expected; +1 moderate positive effect expected; 0 no effect or neutral impact expected; -1 moderate negative effect expected; -2 major negative effect expected.

Table 26– Comparison of impacts of reduced rates for small cider makers

Impact area and No change Introducing reduced rates to small cider target groups makers

Tax revenues for Impacts are estimated as negligible in public authorities Reduced rates cannot be granted to most of the sample MS, and modest in 0 small producers of OFB, and thus 0/ -1 Ireland and the UK. Total foregone no costs arise for public budgets. revenues at EU level estimated at about EUR 15 mn.

164

Impact area and No change Introducing reduced rates to small cider target groups makers

Market effects for The regulatory framework for Small cider makers would gain economic operators alcoholic beverages foresees relatively to large producers. Market 0 different treatment for different +1 effects are estimated to remain small, producers. However, the situation given the limited amount of sales is not expected to change. covered by the reduction.

Health impacts for At EU level, the amount of alcoholic consumers As there are no reduced rates for beverages concerned is very limited. 0 OFB, per capita alcohol 0 / -1 Impacts could be noticeable only in MS consumption is not affected. with a large cider market (such as UK and Ireland).

SME competitive- The competitiveness of small cider ness Current competitiveness of small makers would be greatly enhanced by 0 producers will remain unchanged if +2 the provision. Diseconomies of scale no intervention is brought forward. and market access barriers could be counterbalanced.

Administrative The lack of reduced rates for OFB Administrative burdens for beneficiaries burdens for economic 0 generates no administrative 0 / -1 from reduced rates are estimated to be operators burdens. negligible, at 0.32 €/hl.

Enforcement costs The number of economic operators for public authorities The lack of reduced rates for OFB, concerned, the amount of excise 0 generates no enforcement costs for 0/ -1 revenues at stake, and the marginal role public authorities. of cross-border trade would not require significant additional resources. Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017. Legend: +2 major positive effect expected; +1 moderate positive effect expected; 0 no effect or neutral impact expected; -1 moderate negative effect expected; -2 major negative effect expected.

Table 27 – Comparison of impacts on increasing the threshold of reduced rates for low-strength beer Impact area No Change Revised threshold for low strength beer

In MS opting for the new provision, an MS will keep applying the additional share of beer consumed will Tax revenues 0 current structure and excise -1 benefit from reduced rates, thus tax duty rates. revenues would decline. In MS opting for the new provision, low- strength beer will pay lower tax. No change in taxation; Depending on the extent to which the Market effects 0 therefore, no change in price +1 discount is passed on to consumers, its and consumption. price will decline and its consumption will increase. Public health impacts are negligible due to limited consumption growth in No change in consumption, absolute terms (up to 0.1 L per capita of Public health 0 therefore, no impacts on -1 additional beer consumption) but some public health. population groups e.g. youth, (pregnant) women could be adversely affected. Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017. Legend: +2 major positive effect expected; +1 moderate positive effect expected; 0 no effect or neutral impact expected; -1 moderate negative effect expected; -2 major negative effect expected.

Unclear provisions to measure of Plato degree for sweetened / flavoured beer

165

Table 28 – Comparison of impacts of alternative methods for measuring the Plato degree of sweetened / flavoured beer No Change/ Impact area Option 1.A/2.A Option 1.B.1/2.B.1 Option 1.B.2/2.B.2 baseline Lower Plato Lower Plato degree degree for Higher Plato degree for for MS will keep sweetened/flavour sweetened/flavoured sweetened/flavoured applying their ed beer. beer. Tax revenues 0 -1 -1 beer. 0 preferred Most MS will Few MS will have to Most MS will have approach. have to change change their current to change their their current approach. current approach. approach. In most MS, In most MS, sweetened/flavour In few MS, sweetened / ed beer will pay sweetened/flavoured flavoured beer will No change in lower tax. beer will pay higher pay lower tax. taxation; Depending on the tax. Depending on the Depending on the Economic therefore, no extent to which 0 /- extent to which the 0 +1 +1 extent to which the effects change in the discount is 1 additional tax is passed discount is passed price and passed on to on to consumers, price on to consumers, consumption. consumers, price will increase and price will decline will decline and consumption will and consumption consumption will decrease. will increase. increase. Possible disparities in Some the level of taxation of competition No relevant sweetened/flavoured distortions in disparities in the No relevant beer and standard beer MS applying level of taxation disparities in the of the same alcohol approach B2 of level of taxation of strength. Since most of are possible, 0 /- Market effects -1 +1 sweetened/flavour +1 sweetened/flavoured MS already adopt and may grow 1 ed beer and beer and standard approach B2 and the with the standard beer of beer of the same market of these growth of the same alcohol alcohol strength. products is small, the sweetened/fla strength. overall impact on voured beer market functioning market. would be modest. Public health Public health impacts are Public health impacts impacts are No change in negligible due to are negligible due to Public health 0 0 0 negligible due to 0 consumption. limited limited consumption limited consumption consumption reduction. growth. growth. Most MS will have to implement Most MS will have new enforcement to implement new procedures to test enforcement the Plato degree All MS will be able to procedures to test MS will keep of measure the Plato the Plato degree of Enforcement applying sweetened/flavour degree of sweetened/flavoured costs 0 current -1 ed beer and -1 +1 sweetened/flavoured beer and perform enforcement perform on-site beer based on the on-site checks. procedures. checks. analysis of the final Coordination at the Coordination at product EU level required the EU level for beer moved required for beer across borders. moved across borders. The pending Increased legal Increased legal Increased legal Litigation CJEU certainty (yet, in certainty (yet, in the certainty (yet, in the +1 +1 +1 +1 costs judgment (C- the short-run non- short-run non- short-run non-binding 30/17) may binding guidelines binding guidelines guidelines may increase 166

eventually may increase may increase litigation costs). shed light on litigation costs). litigation costs). the correct interpretation of the terms ‘finished product’. Source: Economisti Associati, "Study on Council Directive 92/83/EEC on the structures of excise duty on alcohol and alcoholic beverages", 2017. Legend: +2 major positive effect expected; +1 moderate positive effect expected; 0 no effect or neutral impact expected; -1 moderate negative effect expected; -2 major negative effect expected.

167

ANNEX 18. MONITORING AND EVALUATION OF IMPACTS

Expected Problem M result/impact Indicators (examples) E Data sources/frequency134 area 133 s

Dysfunctions increased gradual adoption of Commission excise in the legal Eurodenaturant by all MS statistics/analysis (S/R) application certainty for or full recognition of all  of the economic notified national denatured operators denaturation methods alcohol number of instances of Complaints (O) non-recognition of Infringement cases (O)  denaturing methods of Works of the ExComm other MS Committee (R)

Complaints (O) number of instances of Infringement cases (O) diverging application as to  the indirect use of PDA Works of the ExComm Committee (R)

reduced number of instances of Complaints (O) distortion of non-recognition of Infringement cases (O)  competition/ denaturing methods of market Works of the ExComm other MS Committee (R) barriers Complaints (O) number of instances of Infringement cases (O) diverging application as to  the indirect use of PDA Works of the ExComm Committee (R)

more equal treatment of Market/sector analysis (E)  goods containing PDA Questionnaires to/interviews with industry (E) impact on Market/sector analysis (E) costs savings for users of regulatory PDA in MS that currently Questionnaires to/interviews with costs and industry (E) do not exempt indirect burdens  uses or through lower risks related to of delays and associated recognition costs process

contribution reduced instances and/or Market/sector analysis (E)  to reported amounts of fraud Questionnaire to/ interviews with strengthening and unrecorded excise administrations (E)

133 Monitoring will be done in the context of works of the Committee on Excise Duties. 134 O – ongoing; E – evaluation, at the earliest every 5 years; R – at regular intervals, e.g. annual or bi-annual depending on the nature of the data or the data collection method; S – statistics;

168

the fight consumption against fraud and tax reduced scope for Market/sector analysis (E) evasion international Questionnaire to/ interviews with misclassification of PDA  excise administrations (E) Questionnaires to/interviews with industry (E) Expected Problem result/impact Indicators (examples) M E area s

Dysfunctions increased reduced cross-country Commission excise in the legal disparities (e.g. through statistics/analysis (S/R) classification certainty for harmonised definitions of  Works of the ExComm of certain economic certain OFB or common Committee (R) alcoholic operators rules and methods) beverages Complaints (O) number of instances of Infringement cases (O)  misclassified products Works of the ExComm Committee (R)

reduced Market/sector analysis (E) reduced instances of distortion of Questionnaire to/interviews with 'classification shopping' for competition/  excise administrations (E) market more favourable tax Questionnaires to/interviews with barriers classification industry (E)

impact on reduced (legal) costs of Questionnaires to/interviews with regulatory misclassified OFB  industry (E) costs and products burdens from reclassificatio Questionnaire to/ interviews with costs for adapting existing excise administrations (E) n of OFBs (IT) systems and  implementing new rules Questionnaires to/interviews with industry (E)

impact on tax amount of excise duties Statistics (S) revenues collected Questionnaire to/ interviews with from  excise administrations (E) reclassificatio n of OFBs

impact on prices of and demand for Market/sector analysis (E) market size reclassified OFBs  Questionnaires to/interviews with and trends for industry (E) OFBs Statistics (S) market structure for OFB Market/sector analysis (E) products and for non-target Questionnaires to/interviews with products (reclassified  industry (E) products, withdrawn Statistics (S) products, novel products, etc.);

169

Expected Problem result/impact Indicators (examples) M E area s

Dysfunction increased Questionnaire to/ interviews with al legal reduced cross-country excise administrations (E) application certainty for disparities and recognition  Questionnaires to/interviews with of reduced small of small brewers across the industry (E) EU rates producers Works of the ExComm Committee (R)

number of instances of Complaints (O) CJEU cases on the  Infringement cases (O) interpretation of eligibility criteria

reduced improved competitiveness Market/sector analysis (E) distortion of of the small producers (e.g. Questionnaires to/interviews with competition/ off-setting production  industry (E) market costs and diseconomies of Statistics (S) barriers/ scale for small producers, market especially of cider) discriminatio n for small number of reported Market/sector analysis (E) producers instances of market Questionnaires to/interviews with distortion with respect to industry (E) competitive disruptions or  Statistics (S) unfair tax competition as a result of unavailability of reduced rates

results of the analysis of Market/sector analysis (E) impact on market size and  Questionnaires to/interviews with trends for alcoholic industry (E) beverages Statistics (S) impact on reduced costs (legally and Market/sector analysis (E) regulatory economically independent)  Questionnaires to/interviews with costs and for small producers industry (E) burdens from application of facilitated eligibility Questionnaire to/ interviews with reduced rates controls (enforcement excise administrations (E) costs) for tax  administrations, including in cross-border aspects

regulatory costs linked to Questionnaire to/ interviews with implementation and excise administrations (E)  enforcement of Questionnaires to/interviews with certification industry (E)

impact on tax amount of excise duties Statistics (S)  revenues collected and forgone Questionnaire to/ interviews with excise administrations (E)

ratio of excise duties  Commission analysis (E) collected to regulatory Questionnaire to/ interviews with 170

costs (or perception excise administrations (E) thereof)

impact on Market/sector analysis (E) prices of and demand for market size products enjoying reduced   Questionnaires to/interviews with and trends for rates and other industry (E) alcoholic Statistics (S) beverages market structure for Market/sector analysis (E) products enjoying reduced  Questionnaires to/interviews with rates and other, including industry (E) cross-border trade Statistics (S) substitution effect and Market/sector analysis (E) deflection in consumption Questionnaires to/interviews with patterns (particularly with industry (E)  regards to switching to lower/higher alcoholic drinks)

Market/sector analysis (E) economic relevance of Questionnaires to/interviews with  thresholds industry (E) Questionnaire to/ interviews with excise administrations (E) Expected Problem result/impact Indicators (examples) M E area s

Unclear increased Questionnaire to/ interviews with provisions to legal excise administrations (E) measure certainty for reduced cross-country Questionnaires to/interviews with  Plato degree producers of disparities industry (E) sweetened for Works of the ExComm sweetened / and flavoured Committee (R) flavoured beer beer number of instances of Infringements (O) CJEU cases on the  interpretation of calculation method

reduced number of reported Market/sector analysis (E)

distortion of instances of market Questionnaires to/interviews with competition distortion with respect to industry (E) between competitive disruptions or  Questionnaire to/ interviews with products of unfair tax competition as a excise administrations (E) which the result of the application of alcoholic different methods strength was calculated results of the analysis of Market/sector analysis (E) based on impact on market size and Questionnaires to/interviews with  different trends for sweetened and industry (E) methods flavoured beer Statistics (S)

171

impact on incurred costs (legal, Market/sector analysis (E) administrativ economic) for producers of Questionnaires to/interviews with e costs and sweetened and flavoured  industry (E) burdens from beer from switching to switching to other calculation method another calculation regulatory costs linked to Questionnaire to/ interviews with adaptation of national excise administrations (E) method  monitoring and control Questionnaires to/interviews with systems industry (E) impact on tax amount of excise duties Statistics (S) revenues collected from each type of Questionnaire to/ interviews with product of which the excise administrations (E) alcoholic strength was  calculated based on different methods (including ABV vs Plato) impact on Market/sector analysis (E) prices of and demand for market size Questionnaires to/interviews with sweetened and flavoured   and trends for industry (E) sweetened beer Statistics (S) and flavoured beer Market/sector analysis (E) market structure for Questionnaires to/interviews with sweetened and flavoured  industry (E) beer Statistics (S)

substitution effect and Market/sector analysis (E) deflection in consumption Questionnaires to/interviews with patterns (particularly with  industry (E) regards to switching to other types of alcoholic drinks) relevance of economic relevance, Market/sector analysis (E) Plato/AVB stakeholder perception and Questionnaires to/interviews with differentiatio other determining factors  industry (E) n Questionnaire to/ interviews with excise administrations (E)

172